Swallowing disorder monitoring and intervention equipment, control methods, devices, media and products

Abstract

This application provides a swallowing disorder monitoring and intervention device, control method, equipment, medium, and product, belonging to the field of health monitoring and assisted rehabilitation. The device includes a multimodal sensing component, a stimulation component, and a processing unit. The processing unit is configured to receive multimodal physiological signals in real time; obtain assessment results characterizing the current swallowing function state based on the multimodal physiological signals; generate a combination of intervention stimulation parameters based on the assessment results; and control the stimulation component to output intervention stimulation according to the combination of intervention stimulation parameters. This application deploys multimodal sensing components and stimulation components in the laryngeal and mandibular regions, and uses the processing unit to assess the multimodal physiological signals, generate intervention stimulation parameters, and execute intervention stimulation, forming a real-time closed-loop processing flow. It comprehensively reflects the synergistic relationship between different physiological activities during swallowing through multimodal physiological signals, and also links swallowing function monitoring with rehabilitation intervention in real time, improving the accuracy and adaptability of the intervention.

Description

Technical Field

[0001] This application relates to the fields of health monitoring and assisted rehabilitation technology, and in particular to a swallowing disorder monitoring and intervention device, control method, equipment, medium and product. Background Technology

[0002] Dysphagia is a common complication of many diseases, which seriously affects patients' nutritional intake and quality of life. Therefore, long-term monitoring of swallowing function and auxiliary rehabilitation training are of great significance.

[0003] To meet these needs, existing technologies typically offer solutions in two areas: swallowing signal monitoring and swallowing function intervention. Swallowing signal monitoring usually utilizes sensors placed in the neck to acquire physiological signals related to swallowing activities, enabling the identification of swallowing events or alerts to abnormal states. Swallowing function intervention typically uses surface electrodes to apply specific electrical stimulation to swallowing-related muscle groups, aiming to achieve rehabilitation training through passive muscle contraction. The intervention stimulation parameters of these devices are usually preset fixed parameters or manually adjusted offline by professionals based on periodic assessment results.

[0004] The aforementioned existing technical solutions still have shortcomings in practical applications. First, existing solutions typically rely on a single type of monitoring signal to determine swallowing events, making it difficult to comprehensively reflect the synergistic relationship between different physiological activities during swallowing, resulting in insufficient monitoring accuracy. Second, the functional pathways of swallowing function monitoring and rehabilitation intervention are usually separate, lacking a real-time linkage mechanism. The stimulation program executed by the intervention device cannot be dynamically and instantly adjusted according to the user's functional state in a single swallow, leading to insufficient precision and adaptability of the intervention. Summary of the Invention

[0005] This application provides a swallowing disorder monitoring and intervention device, control method, equipment, medium and product, which aims to solve the problems of insufficient monitoring accuracy due to the single swallowing function monitoring signal, and insufficient precision and adaptability of intervention due to the lack of a real-time linkage mechanism between swallowing function monitoring and rehabilitation intervention.

[0006] In a first aspect, this application provides a swallowing disorder monitoring and intervention device, including a multimodal sensing component, a stimulation component, and a processing unit; The multimodal sensing component is configured to acquire multimodal physiological signals during swallowing in real time; The stimulation component is configured to output interventional stimuli; The processing unit is electrically connected to the multimodal sensing component and the stimulation component, respectively; The processing unit includes a signal receiving module, an evaluation module, a parameter generation module, and a control module; The signal receiving module is configured to receive the multimodal physiological signals in real time; The assessment module is configured to obtain assessment results characterizing the current swallowing function state based on the multimodal physiological signals; The parameter generation module is configured to generate a combination of intervention stimulus parameters based on the evaluation results, and the combination of intervention stimulus parameters includes multiple intervention stimulus parameters. The control module is configured to control the stimulation component to output the intervention stimulus according to the combination of intervention stimulus parameters.

[0007] As one embodiment, the evaluation module includes a score determination module, a grade determination module, and an evaluation result acquisition module; The score determination module is configured to obtain a comprehensive score for swallowing dysfunction and an intervention risk score based on the multimodal physiological signals; The level determination module is configured to determine the risk level based on the intervention risk score, and to determine the assessment level based on the comprehensive score of the swallowing function decline and the intervention risk score; The assessment result acquisition module is configured to use a combination of the swallowing function impairment score, the intervention risk score, the risk level, and the assessment level as the assessment result.

[0008] As one embodiment, the parameter generation module includes a range determination module, a base value determination module, and an adjustment module; The range determination module is configured to determine the upper and lower limits of each of the intervention stimulus parameters corresponding to the assessment level when the risk level is the first target risk level; The baseline value determination module is configured to determine the baseline values ​​of each of the intervention stimulation parameters based on the comprehensive score of the swallowing function impairment. The adjustment module is configured to adjust the base value according to the risk level and the intervention risk score within the parameter range determined by the upper limit and the lower limit, so as to generate the intervention stimulus parameter combination.

[0009] As one embodiment, the multimodal physiological signals include local laryngeal deformation signals, surface electromyography signals, and inertial motion signals; The score determination module includes a feature extraction module, a swallowing function impairment component determination module, and a comprehensive swallowing function impairment score determination module; The feature extraction module is configured to extract electromyographic features from the surface electromyography signal; extract laryngeal local deformation features from the laryngeal local deformation signal; extract inertial response features from the inertial motion signal; determine swallowing response delay features based on the peak time of any one of the laryngeal local deformation signal, surface electromyography signal, and inertial motion signal as a reference; and extract channel coordination features between the multimodal sensing components based on the multimodal physiological signals. The swallowing function impairment component determination module is configured to determine the electromyographic insufficiency component based on the electromyographic characteristics; determine the deformation insufficiency component based on the laryngeal local deformation characteristics; determine the inertial insufficiency component based on the inertial response characteristics; determine the time delay abnormal component based on the swallowing response time delay characteristics; and determine the coordination abnormal component based on the channel coordination characteristics. The swallowing function impairment comprehensive score determination module is configured to perform a weighted summation of the electromyographic insufficiency component, the deformation insufficiency component, the inertia insufficiency component, the time delay abnormality component, and the coordination abnormality component to obtain the swallowing function impairment comprehensive score.

[0010] As one embodiment, the score determination module further includes a risk component determination module and an intervention risk score determination module; The risk component determination module is configured to determine the risk components of abnormal outbreak signal after swallowing, abnormal recovery after swallowing, functional stability fluctuation risk components, and abnormal signal quality risk components based on the multimodal physiological signals. The intervention risk score is configured to be obtained by weighted summation of the risk components of the abnormal outbreak signal after swallowing, the abnormal recovery after swallowing, the functional stability fluctuation risk component, and the abnormal signal quality risk component.

[0011] As one embodiment, the baseline value determination module includes a severity coefficient determination module and a first baseline value acquisition module; The severity coefficient determination module is configured to determine the severity coefficient based on the comprehensive score of the swallowing function impairment; The first base value acquisition module is configured to obtain a base value for at least one of the intervention stimulus parameters based on the severity coefficient, the upper limit value, and the lower limit value of each of the intervention stimulus parameters.

[0012] As one example, the combination of intervention stimulation parameters includes stimulation current intensity and at least one other intervention stimulation parameter; The baseline value determination module further includes a second baseline value acquisition module, which is configured to determine the baseline value of the stimulation current intensity based on the user's personalized current intensity threshold combination, assessment level, and severity coefficient. The personalized current intensity threshold combination includes a sensory current intensity threshold, a motor current intensity threshold, and an upper limit of tolerable current intensity. The sensory current intensity threshold represents the current intensity when the user first perceives the stimulus, the motor current intensity threshold represents the current intensity when the user begins to observe or detect a muscle response to the intervention stimulus, and the upper limit of tolerable current intensity represents the highest current intensity that the user can tolerate. Accordingly, the first base value acquisition module is configured to determine the base values ​​of the other intervention stimulus parameters based on the upper and lower limits of the other intervention stimulus parameters and the severity coefficient.

[0013] As one embodiment, the adjustment module includes a risk correction module, which is configured to determine a risk correction coefficient based on the risk level, and use the risk correction coefficient to correct the base values ​​of each of the intervention stimulus parameters to obtain the combination of intervention stimulus parameters.

[0014] As an example, the evaluation module further includes an anomaly type determination module, which is configured to determine the anomaly type based on the electromyography insufficiency component, the deformation insufficiency component, the time delay anomaly component, the coordination anomaly component, and the functional stability fluctuation risk component. The adjustment module further includes a single-parameter adjustment module, which is configured to, after correcting the base values ​​of each of the intervention stimulus parameters using the risk correction coefficient, determine the target parameter category that needs to be adjusted individually in the combination of intervention stimulus parameters and the adjustment range of the intervention stimulus parameter of the target parameter category according to the anomaly type, and adjust the intervention stimulus parameter of the target parameter category according to the adjustment range to obtain the combination of intervention stimulus parameters.

[0015] As an example, the adjustment module further includes a critical adjustment module, which is configured to reduce each of the intervention stimulus parameters after adjusting the intervention stimulus parameters of the target parameter category according to the abnormality type, when the difference between the intervention risk score and the upper limit of the second target risk level is less than a difference threshold, so as to obtain the combination of intervention stimulus parameters. The second target risk level is higher than the first target risk level.

[0016] As an example, the risk component determination module includes a first risk component determination module, which is configured to determine the electromyographic burst feature value and the inertial response energy within the target time window after swallowing based on the multimodal physiological signal, and to perform a weighted summation of the electromyographic burst feature value and the inertial response energy to obtain the risk component of the abnormal burst signal after swallowing.

[0017] As an example, the risk component determination module includes a second risk component determination module, which is configured to determine the first time required for the local laryngeal deformation signal to recover from the peak to the deformation baseline and the second time required for the inertial motion signal to recover from the peak to the inertial baseline, and to perform a weighted summation of the first time, the second time, and the time delay abnormal component to obtain the post-swallowing recovery abnormal risk component.

[0018] As an example, the risk component determination module includes a third risk component determination module, which is configured to determine the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature, and the swallowing response delay feature, respectively, based on the electromyographic feature, the laryngeal local deformation feature, the inertial response feature, and the swallowing response delay feature, and to perform a weighted summation of the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature fluctuation component, and the swallowing delay fluctuation component to obtain the functional stability fluctuation risk component.

[0019] As an example, the risk component determination module includes a fourth risk component determination module, which is configured to determine motion artifact risk components, channel loss risk components, and electrode contact stability risk components based on the multimodal physiological signals, and to perform a weighted summation of the motion artifact risk components, the channel loss risk components, and the electrode contact stability risk components to obtain the signal quality abnormality risk components.

[0020] Secondly, this application also provides a method for monitoring and intervening in swallowing disorders, including: Real-time reception of multimodal physiological signals collected by multimodal sensing components during swallowing; An assessment result characterizing the current swallowing function status is obtained based on the multimodal physiological signals. Based on the evaluation results, a combination of intervention stimulus parameters is generated, which includes multiple intervention stimulus parameters. Stimulation control instructions are generated based on the combination of intervention stimulation parameters, and the stimulation control instructions are sent to the stimulation component to drive the stimulation component to output intervention stimulation.

[0021] Thirdly, this application also provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-mentioned methods for monitoring and intervening in swallowing disorders.

[0022] Fourthly, this application also provides a non-transitory computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-mentioned methods for monitoring and intervening in swallowing disorders.

[0023] Fifthly, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements any of the above-described methods for monitoring and intervening in swallowing disorders.

[0024] This application provides a swallowing disorder monitoring and intervention device, control method, equipment, medium, and product. Multimodal sensing components and stimulation components are deployed in the larynx and mandibular regions. After evaluating the multimodal physiological signals using a processing unit, the evaluation results are converted into corresponding intervention stimulation parameters, and intervention stimulation is executed, forming a real-time closed-loop processing flow from signal acquisition to intervention stimulation. This not only comprehensively reflects the synergistic relationship between different physiological activities during swallowing through multimodal physiological signals, improving the accuracy of the evaluation, but also links swallowing function monitoring with rehabilitation intervention in real time, improving the precision and adaptability of the intervention. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the principle of the swallowing disorder monitoring and intervention device provided in this application; Figure 2 This is a schematic diagram showing the swallowing disorder monitoring and intervention device provided in this application being worn on the front of the neck and mandibular region of the human body; Figure 3 This is a schematic diagram of the overall structure of the swallowing disorder monitoring and intervention device provided in this application; Figure 4 This is an exploded structural diagram of the laryngeal sensing area of ​​the swallowing disorder monitoring and intervention device provided in this application; Figure 5This is a schematic diagram of the mandibular sensing area of ​​the swallowing disorder monitoring and intervention device provided in this application; Figure 6 This is a flowchart illustrating the swallowing disorder monitoring and intervention control method provided in this application; Figure 7 This is a schematic diagram of the structure of the electronic device provided in this application.

[0027] Figure label: 100. Multimodal sensing component; 200. Stimulation component; 300. Processing unit; 310. Signal receiving module; 320. Evaluation module; 330. Parameter generation module; 340. Control module; 210. Wearable substrate; 211. Connection area; 220. Mandibular sensing area; 221. Second nonwoven substrate; 222. Second hydrogel electrode clip; 223. Third lead wire; 224. Second nonwoven shell; 230. Laryngeal sensing area; 231. First nonwoven fabric substrate; 232. Flexible strain sensor; 233. Nonwoven fabric insulating layer; 234. Triaxial accelerometer; 235. First nonwoven fabric shell; 236. First hydrogel electrode buckle; 240. Detachable connection interface; 250. First microcontroller; 260. Second microcontroller; 270. First wire; 280. Second wire; 710. Processor; 720. Communication interface; 730. Memory; 740. Communication bus. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. The terms "upper," "lower," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are used only for the convenience of describing this application and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Unless otherwise expressly specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly, for example, as a fixed connection, a detachable connection, or an integral connection; a mechanical connection or an electrical connection; a direct connection or an indirect connection through an intermediate medium; or a connection within two elements. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0030] The terms "first," "second," etc., used in this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, a first object can be one or more. Furthermore, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0031] The swallowing disorder monitoring and intervention devices, control methods, equipment, media, and products provided in this application are described below with reference to the accompanying drawings.

[0032] It should be noted that the swallowing disorder monitoring and intervention device provided in this application can collect swallowing-related multimodal physiological signals during swallowing activity detection and determine and implement intervention stimulation programs in real time. Furthermore, this application can be used for early screening of high-risk groups for swallowing disorders, quantitative analysis of whether training movements are in place and whether compensation has occurred, trend follow-up, and tracking of rehabilitation effects. Its application scenarios are not limited to hospital examinations; it can also serve as a supplementary tool to clinical examination methods for continuous monitoring and low-burden assessment in community, elderly care, and home settings.

[0033] like Figure 1As shown, the swallowing disorder monitoring and intervention device provided in this application includes a multimodal sensing component 100, a stimulation component 200, and a processing unit 300. The multimodal sensing component 100 is configured to acquire multimodal physiological signals during swallowing in real time; the stimulation component 200 is configured to output intervention stimuli; the processing unit 300 is electrically connected to both the multimodal sensing component and the stimulation component, enabling the processing unit to receive data acquired from the multimodal sensing component and send control commands to the stimulation component.

[0034] The processing unit 300 includes a signal receiving module 310, an evaluation module 320, a parameter generation module 330, and a control module 340. The signal receiving module 310 is configured to receive the multimodal physiological signals in real time; The assessment module 320 is configured to obtain an assessment result characterizing the current swallowing function state based on the multimodal physiological signals; The parameter generation module 330 is configured to generate a combination of intervention stimulus parameters based on the evaluation results, the combination of intervention stimulus parameters including multiple intervention stimulus parameters; The control module 340 is configured to control the stimulation component to output the intervention stimulus according to the combination of intervention stimulus parameters.

[0035] Figures 2 to 5 An embodiment of the swallowing disorder monitoring and intervention device provided in this application is given. Figure 2 and Figure 3 As shown, the swallowing disorder monitoring and intervention device has a structure similar to a neck warmer, including a wearable base 210, a mandibular sensing area 220 disposed outside the wearable base 210, and a laryngeal sensing area 230 disposed on the wearable base 210. The mandibular sensing area 220 is electrically connected to the wearable base 210 via a detachable connection interface 240. The wearable base 210 also has a first microcontroller 250 and a second microcontroller 260. The laryngeal sensing area 230 is electrically connected to the first microcontroller 250 via a first wire 270 and to the second microcontroller 260 via a second wire 280. The first microcontroller 250 is electrically connected to the mandibular sensing area 220 via the detachable connection interface 240. Understandably, the second microcontroller 260 can also be electrically connected to the mandibular sensing area 220 via the detachable connection interface 240. The swallowing disorder monitoring and intervention device also includes a host computer, which is connected to the first microcontroller 250 and the second microcontroller 260 for communication.

[0036] like Figure 3As shown, the wearable base 210 is made of a flexible fabric material, such as non-woven fabric, elastic fabric, or composite fabric, to balance skin comfort, structural load-bearing capacity, and long-term wearing stability. The wearable base 210 is elongated in shape, with connecting areas 211 at both ends. These two connecting areas 211 are used to connect to each other around the user's neck for wearing. The connecting areas 211 can use any of the following: Velcro, snaps, magnetic connectors, buckles, or elastic clasps, to accommodate different neck circumferences and ensure wearing stability.

[0037] A laryngeal sensing area 230 is located in the middle of the wearable base 210, corresponding to the area near the user's Adam's apple, thyroid cartilage, and cricoid cartilage, and is used to acquire laryngeal physiological signals during swallowing. A first microcontroller 250 is located on one side of the laryngeal sensing area 230 and is used to complete multimodal signal acquisition, preprocessing, synchronization, and transmission of multimodal signals to the host computer. A second microcontroller 260 is located on the other side of the laryngeal sensing area 230 and is used to complete electrical stimulation intervention control. This layout can reduce the impact on the flexible fitting area in the middle of the larynx and facilitate the acquisition, transmission, and intervention control. A first lead wire 270 and a second lead wire 280 are arranged inside the wearable base 210 and are used to connect the laryngeal sensing area 230, the first microcontroller 250, the second microcontroller 260, and the mandibular sensing area 220. The mandibular sensing area 220 is disposed on the outside of the wearable substrate 210 and connected to the wearable substrate 210 via a detachable connection interface 240. The mandibular sensing area 220 is used to conform to the area below the mandible and collect relevant surface electrical signals, thereby realizing joint monitoring and intervention stimulation of the laryngeal and mandibular regions. The detachable connection interface 240 can be any one of a plug-in interface, a snap-on interface, a magnetic interface, or a flexible ribbon cable interface.

[0038] like Figure 4 As shown, the throat sensing area 230 includes a first nonwoven fabric substrate 231, a flexible strain sensor 232, a nonwoven fabric insulating layer 233, a triaxial accelerometer 234, a first nonwoven fabric shell 235, and a first hydrogel electrode buckle 236.

[0039] The first nonwoven fabric substrate 231 serves as the skin-contact bearing layer, contacting the skin of the anterior neck and supporting the flexible strain sensor 232. The flexible strain sensor 232, mounted on the first nonwoven fabric substrate 231, senses the localized stretching, bulging, and displacement of the skin surface and soft tissues of the larynx during swallowing. A nonwoven fabric insulating layer 233 is disposed outside the flexible strain sensor 232 for insulation and interlayer structural stability. A triaxial accelerometer 234 is positioned between the nonwoven fabric insulating layer 233 and the first nonwoven fabric outer shell 235, monitoring vibration, acceleration, and posture changes in the laryngeal region. The first nonwoven fabric outer shell 235, located on the outermost layer, forms a protective layer and structural encapsulation.

[0040] In one possible implementation, a flexible strain sensor 232 is used to collect local deformation signals in the larynx. The flexible strain sensor 232 employs a dual-array structure, with two sensing areas covering the laryngeal protuberance region and the cricoid cartilage region, respectively, to improve the detection capability of larynx elevation, forward movement, and local deformation during swallowing. A first hydrogel electrode clip 236 is positioned at the corresponding location in the larynx sensing area and is exposed to the skin through the multi-layer structure for installation of the hydrogel electrode. The first hydrogel electrode clip 236, the flexible strain sensor 232, and the triaxial accelerometer 234 are electrically connected to a first microcontroller 250 via a first wire 270. The first hydrogel electrode clip 236 is electrically connected to a second microcontroller 260 via a second wire 280. The installed hydrogel electrode can be used as a surface electromyography (EMG) acquisition electrode to collect surface EMG signals and as a stimulation output electrode to output interventional stimulation.

[0041] The flexible strain sensor 232 can be fabricated using screen printing, inkjet printing, conductive fabric attachment, flexible circuit technology, liquid metal conductive structure, or nano-conductive composite material.

[0042] like Figure 5 As shown, the mandibular sensing area 220 includes a second nonwoven fabric substrate 221, a second hydrogel electrode clip 222, a third wire 223, and a second nonwoven fabric shell 224.

[0043] The second nonwoven fabric substrate 221 is a flexible triangular or fan-shaped sheet conforming to the mandibular curve. Multiple second hydrogel electrode clips 222 are disposed on the surface of the second nonwoven fabric substrate 221 for mounting hydrogel electrodes, which collect surface electrical signals from the area below the mandible and / or the suprahyoid muscle group. A third wire 223 connects the various second hydrogel electrode clips 222 to a detachable connection interface 240, which is electrically connected to the first microcontroller 250 and the second microcontroller 260 on the wearable substrate 210. This detachable structure allows the mandibular sensing area 220 to be removed for cleaning, replacement, consumable replacement, and local maintenance, and also facilitates position adjustment according to different users' mandibular contours and attachment locations. For users with different neck circumferences, mandibular shapes, and attachment requirements, this significantly improves practical usability and engineering adaptability. A second nonwoven fabric outer shell 224 is disposed on the outside, forming a protective layer for the second hydrogel electrode clips 222 and improving overall structural stability.

[0044] The number of hydrogel electrode clips in the laryngeal sensing area 230 and the mandibular sensing area 220 can be set to 2, 4, 6 or more. The hydrogel electrodes can be installed at the corresponding hydrogel electrode clip positions using a detachable, snap-on, magnetic, or flexible cable connection structure.

[0045] As can be seen from the above, in Figures 2 to 5In the illustrated embodiment, the multimodal sensing component 100 includes a flexible strain sensor 232, a triaxial accelerometer 234, and multiple hydrogel electrodes disposed in the laryngeal sensing area 230 and the mandibular sensing area 220. In this way, the multimodal sensing component can acquire multiple physiological signals from the laryngeal and mandibular regions related to swallowing. The stimulation component 200 includes multiple hydrogel electrodes disposed in the laryngeal sensing area 230 and the mandibular sensing area 220 to achieve interventional stimulation of the laryngeal and mandibular regions. The processing unit 300 includes a host computer, a first microcontroller 250, and a second microcontroller 260, wherein a signal receiving module 310 is disposed on the first microcontroller 250, an evaluation module 320 and a parameter generation module 330 are disposed on the host computer, and a control module 340 is disposed on the second microcontroller 260.

[0046] Based on the above embodiments, it is understood that the triaxial accelerometer can be replaced with a higher-dimensional inertial measurement unit. The hydrogel electrode can be replaced with conductive silicone electrodes, flexible fabric electrodes, silver / silver chloride electrodes, or other biocompatible electrodes, which can achieve signal acquisition and intervention stimulation. The first and second microcontrollers can also be integrated into one unit or integrated into a host computer as a single processing unit. A battery, wireless communication module, local storage module, status indicator module, button module, vibration alert module, or charging management module can also be integrated onto the wearable substrate. In addition to the flexible strain sensor and triaxial accelerometer, the laryngeal sensing area can also integrate acoustic sensors, pressure sensors, temperature sensors, humidity sensors, or bioimpedance detection modules.

[0047] This application embodiment deploys physiological signal acquisition sensors and intervention stimulation elements in both the anterior larynx region and the mandibular region. Flexible strain sensors, triaxial accelerometers, and hydrogel electrodes are also deployed in these two regions to acquire multimodal physiological signals, including local laryngeal deformation signals, surface electromyography signals, and inertial motion signals, thereby comprehensively reflecting the synergistic relationship between different physiological activities during swallowing. Functionally, this application embodiment adopts a neckband-style integrated flexible wearable structure, integrating the laryngeal sensing area, mandibular sensing area, acquisition and transmission module, and stimulation control module into a single device. This avoids the problems of complex layout, difficulty in repeated application, and high usage threshold caused by multiple independent patches, independent leads, and distributed modules in existing solutions, and is also more conducive to repeated use in home, community, and clinical settings.

[0048] The processing unit 300 will be described in detail below using the above embodiment as an example.

[0049] Specifically, the first microcontroller 250 synchronously acquires signals from each sensor and aligns the clock via the signal receiving module 310. The acquired data is then preprocessed, including filtering, buffering, encapsulation, and preliminary organization, to form multimodal physiological signals, which are then uploaded to the host computer via the communication module. The host computer's evaluation module 320 analyzes and processes the received multimodal physiological signals, including analyzing signal amplitude, timing relationships, energy distribution, and other characteristics to determine if there are problems such as insufficient laryngeal elevation, delayed initiation, or poor muscle coordination during the current swallowing. This results in a conclusion that can quantitatively or qualitatively describe the quality, efficiency, or risk of the user's current swallowing action—an evaluation result characterizing the current state of swallowing function.

[0050] When acquiring multimodal physiological signals, for the flexible strain sensor 232, a voltage divider sampling method can be used to obtain its resistance change. If the external reference resistor is... The power supply voltage is The sampling node voltage is The equivalent resistance of the flexible strain sensor It can be represented as: (1).

[0051] By measuring the equivalent resistance By performing continuous sampling, the dynamic change curve of resistance during the local deformation process of the throat can be obtained, which can be used to reflect the local displacement and traction changes of the throat.

[0052] Furthermore, if at time The collected sensor resistance value is The average resistance in the resting state As a benchmark, the relative rate of change is commonly used in implementation. It can be represented as: (2); in, It is a dimensionless characteristic quantity that characterizes the intensity of local deformation in the larynx during swallowing, serving as a signal of local deformation in the larynx.

[0053] For surface electromyography (EMG) signals, differential acquisition can be used. If the outputs of the two sampling electrodes are respectively... and Then the surface electromyography signal It can be represented as: (3).

[0054] Surface electromyography (EMG) signals are used to reflect the activation intensity and temporal relationship of related muscle groups.

[0055] To enhance the electromyographic envelope characteristics, the root mean square value of the sliding window can be further calculated to characterize the activation intensity of swallowing-related muscle groups within the corresponding time period. Let the window length be... Then the first Root mean square characteristics of electromyography within a window for: (4); in, It is the squared value of the nth electromyographic characteristic fluctuation component.

[0056] For a triaxial accelerometer, the composite acceleration amplitude can be used to characterize the intensity of laryngeal mechanical vibration, serving as an inertial motion signal to reflect vibration and motion changes during swallowing. If the triaxial accelerations are respectively... , , Then the combined acceleration amplitude for: (5); By analyzing the composite acceleration amplitude The peak value, duration, or local energy change can be used to extract vibration response characteristics in swallowing events.

[0057] In the preprocessing stage, to reduce high-frequency noise and power frequency interference, a combination of bandpass filtering and notch filtering can be used to process the surface electromyography (EMG) signal. Let the original EMG signal be... The filtered signal It can be represented as: (6); in, This indicates bandpass filtering. This indicates power frequency notch filtering.

[0058] For flexible strain and acceleration signals, the mean of a sliding window can be calculated to improve the stability of event detection. If any channel discrete signal is... Window length is Then the moving average output for: (7); Furthermore, to achieve multimodal synchronization and alignment, it is assumed that the three modalities are timestamped at the acquisition end to form separate sequences. , , In the same time window Internal construction of joint observation matrix : (8); in, This represents a sequence of signals indicating laryngeal deformation. Represents surface electromyography sequences. This represents a sequence of inertial motions.

[0059] Through joint observation matrix This allows for the extraction of multimodal joint features within the same time window to obtain joint energy features. Joint energy features within a time window of length T. Represented as: (9) in, Indicates the first The mode in the th ... The amplitude of each sampling point.

[0060] Based on joint energy characteristics A sliding window analysis is performed on continuously acquired multimodal joint features to determine whether a swallowing event has occurred and to define its time boundaries. Specifically, baseline signals are first acquired when the user is in a resting or non-swallowing state, and event trigger thresholds, start thresholds, and end thresholds are determined based on these baseline signals. When the joint energy feature exceeds the preset event trigger threshold within multiple consecutive detection windows, and at least two of the following signals—laryngeal local deformation signal, surface electromyography signal, and inertial motion signal—show corresponding responses within a preset time range, a candidate swallowing event is determined to have occurred. If only a single modality shows a transient high-amplitude response, while other modalities do not show corresponding changes, the response can be judged as a non-swallowing artifact such as head and neck movement, coughing, speaking, abnormal electrode contact, or external collision, and is not considered a valid swallowing event.

[0061] After determining the occurrence of a candidate swallowing event, the event boundary is further determined based on the changes in the combined energy feature relative to the start and end thresholds. Specifically, the moment when the combined energy feature first exceeds a preset start threshold before the event triggering time and meets the start confirmation condition within a subsequent detection window is determined as the start time of the swallowing event. The start confirmation condition includes at least one of the following: the combined energy feature remains above a preset maintenance threshold within the subsequent detection window; the rising slope of the combined energy feature within the subsequent detection window meets a preset slope condition; at least two of the laryngeal local deformation signal, surface electrical signal, and inertial signal meet their respective response threshold conditions within a preset synchronization time range; and the signal quality anomaly index within the subsequent detection window is below a preset anomaly threshold. The moment when the combined energy feature falls below a preset end threshold after the event triggering time and remains below this end threshold for multiple consecutive detection windows is determined as the end time of the swallowing event. This allows for the division of swallowing event time windows with clearly defined start and end times from continuously acquired multimodal signals, within which subsequent electromyographic features, laryngeal deformation features, inertial response features, and swallowing timing features can be extracted.

[0062] The above process enables the synchronous acquisition and unified output of local laryngeal deformation signals, surface electromyography signals, and inertial motion signals, providing basic data for subsequent swallowing function recognition, training feedback, and trend analysis.

[0063] The parameter generation module 330 receives the evaluation results from the evaluation module 320 and analyzes them according to preset logical rules to determine whether intervention stimulation is needed. If intervention stimulation is needed, the evaluation results are mapped to a corresponding set of personalized intervention stimulation parameter combinations. The intervention stimulation is a physical stimulus applied to swallowing-related muscle groups to assist rehabilitation training, such as electrical stimulation output from hydrogel electrodes. The intervention stimulation parameter combination is a set of specific parameters used to precisely define the characteristics of the intervention stimulation and may include multiple intervention stimulation parameters, such as stimulation frequency, pulse width, and current intensity.

[0064] Intervention Stimulus Parameter Combination It can be represented as: (10); in, Set the function for the parameters.

[0065] Preferably, the parameter setting function At least the following rules must be met: a. Assessment Level Used to determine the baseline range of stimulus parameters.

[0066] b. Overall score for impaired swallowing function Used to determine the specific value position within the same evaluation level.

[0067] c. Risk Level Used to determine whether stimulus output needs to be reduced.

[0068] d. Intervention risk score Used to impose continuous constraints on stimulus parameters.

[0069] Therefore, parameter setting function Stimulus parameters are not determined solely based on a single assessment level, but rather according to the assessment level. Risk level Comprehensive score for impaired swallowing function Intervention risk score Determine the stimulus parameters so that different assessment results can correspond to different combinations of intervention stimulus parameters.

[0070] After the parameter generation module 330 generates a specific combination of intervention stimulus parameters, it transmits this combination to the second microcontroller 260 in the form of a control command U. The control command may include specific stimulus parameters, start flags, and stop flags. It can be represented as: (11); in, For the combination of intervention stimulus parameters, As a stimulus activation marker, This is a sign that the stimulus has stopped.

[0071] After receiving the control command, the control module 340 in the second microcontroller 260 converts the set of parameters into stimulation control instructions that can be directly executed by the hardware circuit of the stimulation component. The stimulation control instructions are used to control the stimulation output module to generate electrical stimulation signals with predetermined frequency, pulse width, intensity and timing.

[0072] In one possible implementation, the stimulation control state of the control module 340 It can be represented as: (12); in, This indicates that the output of the stimulus signal is permitted; This indicates that no stimulus signal is output; G represents the evaluation level. The risk level is indicated in the following explanation.

[0073] Subsequently, the control module 340 issues a stimulation control command, driving the hydrogel electrode (stimulation component) to output intervention stimulation signals as required. When the set total stimulation duration is reached or a stop command is received, the control module 340 stops the stimulation output, thereby completing the stimulation intervention.

[0074] Specifically, once the second microcontroller 260 confirms that the output of the stimulation signal is permitted, the stimulation output path outputs an electrical stimulation signal to the hydrogel electrode according to the control command.

[0075] In one possible implementation, the stimulation signal is in the form of a biphasic rectangular pulse to reduce the risk of DC bias and skin polarization. The biphasic pulse may include a positive-phase pulse and a negative-phase pulse, with the positive-phase charge being substantially equal to the negative-phase charge. The charge balance relationship can be expressed as: (13); in, and These represent the current intensity and pulse width of the positive phase pulse, respectively. and These represent the current intensity and pulse width of the negative phase pulse, respectively.

[0076] Understandably, the stimulation signal may also be a symmetrical biphasic pulse, an asymmetrical biphasic pulse, an intermittent pulse train, or other waveforms suitable for electrical stimulation of the body surface.

[0077] During the stimulus output process, when the set total stimulus duration is reached... Upon receiving a stop command, the second microcontroller 260 stops the stimulation output and records the stimulation parameters for this intervention. The recorded content... At least include the evaluation results obtained from evaluation module 320 The parameter generation module 330 generates the combination of intervention stimulus parameters. Stimulation start time and stimulation end time , represented as: (14).

[0078] In this embodiment, multimodal sensing and stimulation components are deployed in the larynx and mandibular regions. After evaluating the multimodal physiological signals using a processing unit, the evaluation results are converted into corresponding intervention stimulation parameters, and intervention stimulation is executed, forming a real-time closed-loop processing flow from signal acquisition to intervention stimulation. This not only comprehensively reflects the synergistic relationship between different physiological activities during swallowing through multimodal physiological signals, improving the accuracy of the evaluation, but also links swallowing function monitoring with rehabilitation intervention in real time, improving the precision and adaptability of the intervention.

[0079] In one possible implementation, the evaluation module 320 includes a score determination module, a grade determination module, and an evaluation result acquisition module.

[0080] The score determination module is configured to obtain a comprehensive score for swallowing function decline and an intervention risk score based on the multimodal physiological signals.

[0081] The scoring module performs in-depth feature extraction and analysis on the received multimodal physiological signals, quantifying them from two dimensions: functional performance and potential risk. The comprehensive score F for swallowing function decline is used to characterize the degree of functional decline in the current swallowing event compared to a healthy or ideal state, while the intervention risk score R is used to characterize the abnormal risks that may accompany the implementation of intervention stimulation in the current state, such as the presence of unstable physiological responses or poor signal quality.

[0082] The level determination module is configured to determine the risk level based on the intervention risk score, and to determine the assessment level based on the comprehensive score of the swallowing function decline and the intervention risk score.

[0083] In one possible implementation, the risk level is determined based on the intervention risk score R. At that time, a first risk threshold can be preset. Second risk threshold , where 0 < < .

[0084] when hour, A rating of 0 indicates a low current risk; when hour, Being classified as Level 1 indicates a certain level of risk; subsequent interventions should be cautious, and the intensity of intervention stimulus parameters should be reduced or subject to manual confirmation. hour, A rating of Level 2 indicates a high risk, and automatic intervention is not recommended.

[0085] Based on the above, risk level Represented as: (15).

[0086] When determining the assessment level G, the overall swallowing function status of the user is comprehensively graded by combining the comprehensive score F of the swallowing function decline and the intervention risk score R. This process can be implemented using a segmented assessment method to improve the accuracy and interpretability of the assessment results.

[0087] A first functional threshold can be set for the comprehensive score F of impaired swallowing function. Second functional threshold and third functional threshold , .

[0088] When segmenting, when and At that time, it is determined that the current swallowing function is basically stable; when ,and At that time, it was determined that there was a mild decline in swallowing function; when and At: A moderate decline in swallowing function is determined; when and At that time, it was determined that there was a significant decline in swallowing function; when If the situation is deemed to be of high risk, it is not recommended to proceed directly to the subsequent stimulus implementation phase.

[0089] To establish the correlation between swallowing function assessment results and subsequent intervention stimulus parameter selection, the current swallowing function status was further discretized into assessment levels. Among them, the assessment level Based on the comprehensive score of impaired swallowing function Intervention risk score To be determined jointly.

[0090] Based on the above, the assessment level It can be represented as: (16).

[0091] In one alternative implementation, , , , and It can be set based on baseline data, historical assessment results, or clinically labeled data of healthy users, and can be individually adjusted according to the age, disease course, basic muscle strength, rehabilitation stage, clinical labeled data, or device sensitivity of different users.

[0092] The assessment result acquisition module is configured to use a combination of the swallowing function impairment score, the intervention risk score, the risk level, and the assessment level as the assessment result.

[0093] Specifically, the comprehensive score of the swallowing function decline, the intervention risk score, the risk level, and the assessment level are combined to output a comprehensive assessment conclusion containing multi-dimensional information.

[0094] Based on the specific risk level and assessment level mentioned above, the parameter setting function... Follow these specific rules: a. Assessment Level Used to determine the baseline range of stimulus parameters. When, no stimulus parameters are output; when , or When, they correspond to parameter ranges from low to high; when When this happens, the stimulus parameters are not output directly, and a prompt is given for manual review or reassessment.

[0095] b. Overall score for impaired swallowing function Used to determine the specific value position within the same evaluation level. When When the level increases within the range corresponding to the current assessment grade, the stimulation dose can be increased within that parameter range. Increasing the stimulation dose may include increasing the stimulation frequency. Increase pulse width Increase current intensity Prolonging the duration of stimulation Or prolong the total stimulation duration At least one of the methods.

[0096] c. Risk Level Used to determine whether stimulus output needs to be reduced. When When, set the parameters according to the parameter range corresponding to the current assessment level; when When this occurs, reduce the current intensity, shorten the total stimulation duration, or extend the stop time within the corresponding parameter range; when When this occurs, no stimulus parameters are output.

[0097] d. Intervention risk score Used to impose continuous constraints on stimulus parameters. When approaching the risk threshold, even if the assessment level If the current is too high, the current intensity also needs to be reduced. Shorten the duration of stimulation Shorten the total stimulation duration Extend the stop time Or reduce the duty cycle This is to avoid delivering excessive stimulation under high-risk conditions.

[0098] Therefore, parameter setting function Stimulus parameters are not determined solely based on a single assessment level, but rather by a comprehensive consideration of all assessment levels. Risk level Main types of anomalies Comprehensive score for impaired swallowing function Intervention risk score and individual current intensity threshold set This leads to different comprehensive evaluation results It can accommodate different combinations of stimulus parameters.

[0099] This application embodiment quantifies the original multimodal physiological signals into multidimensional, structured assessment results by determining the comprehensive score of swallowing function decline and the intervention risk score, so that the assessment conclusions comprehensively reflect multidimensional information.

[0100] In one possible implementation, the multimodal physiological signals include local laryngeal deformation signals, surface electromyography signals, and inertial motion signals, as described above.

[0101] The score determination module includes a feature extraction module, a swallowing function impairment component determination module, and a comprehensive swallowing function impairment score determination module.

[0102] The feature extraction module is configured to extract electromyographic features from the surface electromyography signal; extract laryngeal local deformation features from the laryngeal local deformation signal; extract inertial response features from the inertial motion signal; determine swallowing response delay features based on the peak time of any one of the laryngeal local deformation signal, surface electromyography signal, and inertial motion signal as a reference; and extract channel coordination features between the multimodal sensing components based on the multimodal physiological signals.

[0103] The feature extraction module is configured to extract a set of feature parameters for quantitative evaluation from the multi-dimensional raw signal within the corresponding swallowing event time window after a swallowing event is detected. Specifically, the electromyographic (EMG) features characterize the activation intensity and recruitment degree of swallowing-related muscle groups, including peak EMG envelope, integrated EMG value, and EMG duration. The laryngeal local deformation features characterize the elevation, forward movement, and local traction changes of the larynx during swallowing, including deformation peak, deformation duration, and deformation recovery time. The inertial response features characterize the vibration and motion changes of the laryngeal region during swallowing, including inertial response peak and inertial response energy. The swallowing response delay features characterize the temporal relationship between swallowing initiation, muscle activation, laryngeal elevation, and recovery; the channel coordination features characterize the degree of coordination between different acquisition channels. Each sensing element in the multimodal sensing component includes at least one acquisition channel; in the above embodiment, each hydrogel electrode can be considered as one acquisition channel.

[0104] In one possible implementation, the integrated electromyographic value of the current swallowing event is used as the electromyographic feature. It can be represented as: (17); in, The surface electromyography signal within the swallowing event window. The moment when the swallowing event begins. The moment the swallowing event ends. This represents the integrated electromyographic value (i.e., electromyographic characteristics) of the current swallowing event.

[0105] Local deformation characteristics of the larynx It can be represented as: (18); in, This is a signal of localized laryngeal deformation within the swallowing event window. This is the baseline value for the channel in its resting state. This describes the local deformation characteristics of the larynx during the current swallowing event.

[0106] Inertial response characteristics It can be represented as: (19); in, , , These are the acceleration signals from the triaxial accelerometer in three directions. This represents the inertial response characteristics of the current swallowing event.

[0107] Swallowing response delay characteristics It can be represented as: (20); in, The reference time is any one of the following: the peak activation time of the main muscle groups, the peak deformation time of the larynx, or the peak inertial response time. This represents the current swallowing response delay characteristics.

[0108] In one possible implementation, the channel coordination coefficient As a characteristic of channel coordination, the channel coordination coefficient. The normalized maximum correlation between the two related channels within the swallowing event window can be used to represent this: (twenty one); in, and Feature signals are extracted from the signals of the two related channels respectively. The related channels can be any combination of the left electromyography channel and the right electromyography channel, the laryngeal electromyography channel and the mandibular electromyography channel, the laryngeal deformation channel and the mandibular electromyography channel, etc. This is the amount of time delay. This represents the maximum allowed time offset range.

[0109] The swallowing function impairment component determination module is configured to determine the insufficient electromyographic component based on the electromyographic characteristics; determine the insufficient deformation component based on the local deformation characteristics of the larynx; determine the insufficient inertia component based on the inertial response characteristics; determine the abnormal time delay component based on the swallowing response delay characteristics; and determine the abnormal coordination component based on the channel coordination characteristics.

[0110] Specifically, insufficient electromyographic components It can be represented as: (twenty two); Insufficient deformation component It can be represented as: (twenty three); Insufficient inertial component It can be represented as: (twenty four); Delay anomaly components It can be represented as: (25); Anomalous components of coordination It can be represented as: (26); in, This is the reference value for electromyography. This serves as a reference value for localized deformation of the throat. This is the reference value for the inertial response. This is a reference delay for the swallowing response.

[0111] The swallowing function impairment comprehensive score determination module is configured to perform a weighted summation of the electromyographic insufficiency component, the deformation insufficiency component, the inertia insufficiency component, the time delay abnormality component, and the coordination abnormality component to obtain the swallowing function impairment comprehensive score.

[0112] Specifically, based on the above components, a comprehensive score for impaired swallowing function is constructed. It can be represented as: (27); in, ; , , , and These are the weighting coefficients corresponding to the insufficient electromyography component, insufficient deformation component, insufficient inertia component, abnormal time delay component, and abnormal coordination component, respectively.

[0113] The comprehensive score of impaired swallowing function The larger the value, the more significant the functional decline of the current swallowing event relative to the reference state.

[0114] This application extracts features from multimodal signals that characterize multiple dimensions of muscle activation, structural motion, temporal relationships, and synergy, and unifies them into multiple dimensionless functional decline components, which are then integrated into a comprehensive evaluation score. This solves the problem that existing technologies cannot fully reflect the synergy of the swallowing physiological process, making the evaluation results more objective, comprehensive, and precise.

[0115] In one possible implementation, the score determination module further includes a risk component determination module and an intervention risk score determination module.

[0116] The risk component determination module is configured to determine the risk components of abnormal outbreak signals after swallowing, abnormal recovery after swallowing, functional stability fluctuations, and abnormal signal quality based on the multimodal physiological signals.

[0117] The risk component of abnormal burst signal after swallowing is used to quantify whether there is an abnormal electromyographic burst or inertial vibration within a preset time window after the swallowing action. Such abnormal bursts may be related to adverse reactions such as choking, recurrent laryngeal reflux after aspiration, etc., and are an important indicator for judging the risk of intervention.

[0118] The post-swallowing recovery abnormality risk component is used to assess the time required for swallowing-related physiological indicators (such as laryngeal deformation and electromyographic activity) to recover from peak levels to resting baseline. An excessively long or abnormal recovery process may indicate user fatigue or functional coordination impairment, and proceeding directly to the next stimulus may pose a risk.

[0119] The functional stability fluctuation risk component is used to analyze the consistency of a user's performance across several consecutive swallowing events. It assesses the stability of swallowing function by calculating the degree of fluctuation in core swallowing characteristics (such as peak electromyography (EMG) and deformation amplitude). Large fluctuations indicate an unstable swallowing state, making high-intensity interventions inappropriate.

[0120] The signal quality anomaly risk component is used to assess the reliability of the currently acquired multimodal physiological signals, including the presence of motion artifacts, channel signal loss, poor electrode contact, and other issues. When signal quality is low, the reliability of functional assessments and intervention decisions based on that signal will significantly decrease.

[0121] The intervention risk score determination module is configured to perform a weighted summation of the risk components of the post-swallowing abnormal outbreak signal, the post-swallowing recovery abnormality, the functional stability fluctuation, and the signal quality abnormality to obtain the intervention risk score.

[0122] Intervention risk score It can be represented as: (28); in, This refers to the risk component of abnormal outbreak signals after swallowing. To restore abnormal risk components after swallowing, This represents the component of functional stability fluctuation risk. This is a component representing the risk of signal quality anomalies. , , and These are the weighting coefficients corresponding to each risk component. .

[0123] This application embodiment constructs an independent intervention risk assessment system, which comprehensively assesses the safety and appropriateness of intervention from multiple dimensions such as post-swallowing response, recovery status, functional stability, and signal quality. This avoids potential safety hazards in making intervention decisions based solely on swallowing function assessment, and ensures effective control of the stimulation process when the signal is unreliable or the user's physiological state is at high risk, significantly improving the robustness and safety of the entire closed-loop intervention system.

[0124] In one possible implementation, the risk component determination module includes a first risk component determination module, which is configured to determine the electromyographic burst feature value and the inertial response energy within the target time window after swallowing based on the multimodal physiological signal, and to perform a weighted summation of the electromyographic burst feature value and the inertial response energy to obtain the risk component of the post-swallowing abnormal burst signal.

[0125] In one possible implementation, the risk component of the abnormal burst signal after swallowing. It can be determined based on the intensity of electromyographic bursts and the intensity of inertial vibration bursts within a preset time window after swallowing, and can be expressed as: (29); In the formula, The electromyographic burst characteristic values ​​within a preset time window after swallowing. The baseline value for electromyography is the value after resting or normal swallowing. The inertial response energy within a preset time window after swallowing. This corresponds to the inertial response reference value; and These are the weighting coefficients corresponding to the risks of sudden electromyographic events and sudden inertial events, respectively. .

[0126] The embodiments of this application quantify the electromyographic burst characteristic values ​​and inertial response energy within a preset time window after swallowing to determine whether there is abnormal muscle activity or mechanical vibration after swallowing, providing an objective basis for identifying potential high-risk events such as aspiration or choking.

[0127] In one possible implementation, the risk component determination module includes a second risk component determination module, which is configured to determine the first time required for the local laryngeal deformation signal to recover from its peak to its deformation baseline and the second time required for the inertial motion signal to recover from its peak to its inertial baseline, and to perform a weighted summation of the first time, the second time, and the time delay abnormal component to obtain the post-swallowing recovery abnormal risk component.

[0128] In one possible implementation, the abnormal risk component is recovered after swallowing. It can be determined by combining the recovery time of local laryngeal deformation, the recovery time of inertial response, and the swallowing response delay, and can be expressed as: (30); In the formula, This is the time required for the local deformation signal in the throat to recover from its peak value to a preset baseline range. Reference time for the recovery of local deformation of the larynx; This is the time required for the inertial change signal to recover from its peak value to a preset baseline range. The inertial response recovery reference time; , and These are the weighting coefficients corresponding to deformation recovery anomalies, inertia recovery anomalies, and time delay anomalies, respectively. .

[0129] This application embodiment quantifies the abnormal risk of post-swallowing recovery by measuring the time required for the local deformation signal and inertial change signal of the larynx to recover from the peak to the preset baseline range. It effectively identifies the problem of delayed recovery after swallowing, provides a quantitative basis for judging whether the user has functional fatigue or decreased coordination, avoids starting the next intervention before the physiological state is stable, and ensures the appropriateness of the intervention.

[0130] In one possible implementation, the risk component determination module includes a third risk component determination module, which is configured to determine the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature, and the swallowing response delay feature, respectively, based on the electromyographic feature, the laryngeal local deformation feature, the inertial response feature, and the swallowing response delay feature, and to perform a weighted summation of the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature fluctuation component, and the swallowing delay fluctuation component to obtain the functional stability fluctuation risk component.

[0131] In one possible implementation, the functional stability fluctuation risk component It can be determined based on the fluctuations in electromyographic characteristics, local laryngeal deformation characteristics, inertial response characteristics, and swallowing delay characteristics during several consecutive swallowing events, and can be expressed as: (31); In the formula, The characteristic wave component of electromyography. This represents the fluctuation component of the local deformation characteristics of the throat. The characteristic wave component of the inertial response. This represents the swallowing delay fluctuation component; , , and These are the corresponding weight coefficients. .

[0132] In one possible implementation, in containing Within a statistical window of a swallowing event, the electromyographic characteristic fluctuation component It can be represented as: (32); (33); In the formula, For the first Electromyographic characteristics of a swallowing event To calculate the mean electromyographic characteristics within the statistical window, This is used to count the number of swallowing events within the window.

[0133] Correspondingly, the fluctuation component of the local deformation characteristics of the throat Inertial response characteristic wave component and swallowing delay fluctuation components They can be represented as: (34); (35); (36); in, , and The first Local deformation characteristics of the larynx, inertial response characteristics, and swallowing response delay for each swallowing event; , and These are the mean values ​​of the corresponding features within the statistical window.

[0134] When there are large fluctuations in electromyographic intensity, laryngeal deformation amplitude, inertial response energy, or response delay during several consecutive swallowing events, it indicates poor stability of the current swallowing action. This is the functional stability fluctuation risk component. Increase.

[0135] This application evaluates the repeatability and stability of swallowing actions from a time series perspective. When continuous swallowing shows large fluctuations, the system can promptly identify this unstable state, thereby adopting a more conservative strategy in intervention decisions and avoiding excessive stimulation to users with unstable states.

[0136] In one possible implementation, the risk component determination module includes a fourth risk component determination module, which is configured to determine motion artifact risk components, channel loss risk components, and electrode contact stability risk components based on the multimodal physiological signals, and to perform a weighted summation of the motion artifact risk components, the channel loss risk components, and the electrode contact stability risk components to obtain the signal quality anomaly risk components.

[0137] In one possible implementation, the signal quality anomaly risk component It can be determined based on signal artifact intensity, channel loss, and electrode contact stability, and can be expressed as: (37); In the formula, As a risk component of motion artifacts, To account for the risk of channel loss, This represents the risk component for electrode contact stability. , and These are the corresponding weight coefficients. .

[0138] In one possible implementation, motion artifact risk component It can be represented as: (38); in, This refers to the noise energy within the preset non-swallowing frequency band or abnormally high amplitude fluctuation range. This is a reference value for noise energy.

[0139] In one possible implementation, the channel loss risk component It can be represented as: (39); in, This represents the number of channels with abnormal, saturated, or lost signals during the current assessment period. This represents the total number of channels participating in the evaluation.

[0140] In one possible implementation, the electrode contact stability risk component It can be represented as: (40); in, This represents the equivalent contact resistance between the hydrogel electrode and the skin. This is a preset contact impedance reference value. When no impedance detection circuit is set, the electrode contact stability risk component can also be indirectly characterized by the baseline drift amplitude of the electromyographic signal, the power frequency interference intensity, or the channel saturation ratio.

[0141] This application embodiment utilizes motion artifact risk components, channel loss risk components, and electrode contact stability risk components to construct a complete signal quality assessment mechanism. This mechanism reviews the reliability of the data source before functional assessment and intervention decisions, ensuring that the entire system's decisions are based on high-quality input data. This avoids misjudgments and inappropriate interventions caused by signal distortion, thereby improving the accuracy and safety of assessment and intervention.

[0142] In one possible implementation, the parameter generation module 330 includes a range determination module, a base value determination module, and an adjustment module.

[0143] The range determination module is configured to determine the upper and lower limits of each of the intervention stimulus parameters corresponding to the assessment level when the risk level is the first target risk level.

[0144] In one possible implementation, the combination of intervention stimulus parameters includes stimulus frequency. Pulse width Stimulation current intensity Ascent time Duration of stimulation Stop time Total stimulation duration wait.

[0145] Understandably, the combination of intervention stimulation parameters may also include pulse waveform, pulse polarity, channel sequence, inter-channel delay, or number of stimulation cycles.

[0146] Specifically, the first target risk level corresponds to This means that the current risk is low.

[0147] when , ( When this occurs, it usually indicates mild muscle group under-recruitment, mild local deformation under-recruitment, or mild decreased coordination. In this case, the intervention stimulation parameters should be set within the first parameter range (the lower parameter range). For example, stimulation frequency. Can be set to pulse width Can be set to Current intensity Can be set to Rise time Can be set to Stimulation duration Can be set to Stop time Can be set to Total stimulation duration Can be set to .

[0148] when , ( When this occurs, it usually indicates the simultaneous presence of insufficient electromyography and insufficient local laryngeal deformation, or a significant abnormality in temporal coordination. In such cases, the stimulation parameters should be set within the second parameter range (moderate parameter range). For example, stimulation frequency... Can be set to pulse width Can be set to Current intensity Can be set to Rise time Can be set to duration of stimulation Can be set to Stop time Can be set to Total stimulation duration Can be set to .

[0149] when , ( When the score is 0, it usually indicates a decline in swallowing function. If the levels are high, and the main abnormalities indicate significant muscle recruitment, laryngeal deformity, or a complex functional decline, then the stimulation parameters should be set within the third parameter range (the higher parameter range). For example, stimulation frequency... Can be set to pulse width Can be set to Current intensity Can be set to Rise time Can be set to duration of stimulation Can be set to Stop time Can be set to Total stimulation duration Can be set to .

[0150] when or Approaching the second risk threshold At that time, even if the assessment level If a moderate or significant functional decline is detected, the corresponding higher-level parameter range (e.g., the third parameter range) is not used directly. Instead, the parameters are set within the adjacent lower-level parameter range (e.g., the second parameter range), or the current intensity is reduced, the total stimulation duration is shortened, and the stop time is extended based on the original parameters (e.g., the third parameter range).

[0151] when , At this time, due to certain risks, the third parameter range is not used. Instead, the second parameter range is used, or the current intensity is reduced, the total stimulation duration is shortened, and the stop time is extended based on the third parameter range.

[0152] when or At this time, the system does not output stimulus parameters, but instead outputs a high-risk warning, prompting for re-collection, manual review, or risk management.

[0153] The baseline value determination module is configured to determine the baseline values ​​of each of the intervention stimulation parameters based on the comprehensive score of the swallowing function decline.

[0154] The baseline values ​​of each intervention stimulus parameter are determined by the comprehensive score F of the swallowing function decline. That is, even if two swallowing events have the same assessment level G, the swallow with the higher F score (indicating a more severe decline in swallowing function) will generate a baseline value that is closer to the upper limit of the parameter range, thereby achieving a more precise intensity matching.

[0155] The adjustment module is configured to adjust the base value according to the risk level and the intervention risk score within the parameter range determined by the upper limit and the lower limit, so as to generate the intervention stimulus parameter combination.

[0156] The adjustment module is configured to act as a safety constraint layer within the parameter range determined by the upper and lower limits, based on the risk level. The continuously varying intervention risk score R is used to finally adjust the base values ​​generated above, producing an executable combination of intervention stimulus parameters. For example, when... When this happens, the adjustment module will perform a derating process based on the baseline value, such as reducing the current intensity or shortening the total stimulation duration; when If that happens, then stop the intervention stimulation directly.

[0157] This application embodiment initially determines the range of intervention stimulus parameters based on risk level and assessment level, determines specific baseline values ​​based on the comprehensive score of swallowing function decline, and fine-tunes them within the range based on risk level and intervention risk score. Through a three-step parameter generation mechanism, discrete assessment levels, continuous functional decline scores, and real-time risk assessment results are organically combined, so that the final output intervention stimulus can accurately match the user's functional state and ensure safety, significantly improving the personalization, accuracy, and safety of the intervention.

[0158] In one possible implementation, the base value determination module includes a severity coefficient determination module and a first base value acquisition module.

[0159] The severity coefficient determination module is configured to determine the severity coefficient based on the comprehensive score of the swallowing function impairment.

[0160] To further incorporate the comprehensive score of impaired swallowing function within the same parameter range The impact of defining the severity coefficient within the interval. : when hour: ; when hour: ; when hour: ; Among them, severity coefficient The larger the value, the higher the frequency, the wider the pulse width, the higher the current intensity, or the longer the total stimulation duration can be selected within the corresponding parameter range.

[0161] The first base value acquisition module is configured to obtain a base value for at least one of the intervention stimulus parameters based on the severity coefficient, the upper limit value, and the lower limit value of each of the intervention stimulus parameters.

[0162] In this embodiment of the application, for any intervention stimulus parameter, if the lower limit of the current parameter range is... The upper limit is Then the base value It can be determined by the following formula: (41).

[0163] The embodiments of this application introduce a severity coefficient to provide a clear calculation basis for the fine adjustment process of parameters, ensuring a positive correlation between the comprehensive score of swallowing function impairment and parameter intensity.

[0164] In one possible implementation, the combination of intervention stimulation parameters includes stimulation current intensity and at least one other intervention stimulation parameter.

[0165] The baseline value determination module further includes a second baseline value acquisition module, which is configured to determine the baseline value of the stimulation current intensity based on the user's personalized current intensity threshold combination, the assessment level, and the severity coefficient. The personalized current intensity threshold combination includes a sensory current intensity threshold, a motor current intensity threshold, and a tolerance current intensity upper limit. The sensory current intensity threshold represents the current intensity when the user first perceives the stimulus, the motor current intensity threshold represents the current intensity when the user begins to observe or detect a muscle response to the intervention stimulus, and the tolerance current intensity upper limit represents the highest current intensity that the user can tolerate.

[0166] Accordingly, the first base value acquisition module is configured to determine the base values ​​of the other intervention stimulus parameters based on the upper and lower limits of the other intervention stimulus parameters and the severity coefficient.

[0167] The user's personalized current intensity threshold combination can be measured before the first use of the stimulation intervention function, so that subsequent electrical stimulation parameters can be individually set according to the personalized current intensity threshold combination.

[0168] In one possible implementation, the personalized current intensity threshold combination includes at least a sensed current intensity threshold. Threshold of motion current intensity and the upper limit of withstand current intensity .

[0169] Among them, the sensing current intensity threshold The current intensity corresponding to the user's first perception of the stimulus; motion current intensity threshold. The current intensity corresponding to the observation or detection of a muscle response in the target area; the upper limit of the tolerable current intensity. This represents the highest current intensity that the user can tolerate without experiencing significant discomfort.

[0170] The personalized current intensity threshold combination It can be represented as: (42); satisfy: .

[0171] When measuring personalized current intensity threshold combinations, the stimulation frequency and pulse width are set as preset test parameters. The current intensity is gradually increased from a low level, and the current values ​​at which the user experiences sensory feedback, slight contraction of the target muscle, and maximum tolerable feedback are recorded. The current increment step can be preset according to the device resolution. The aforementioned personalized current intensity threshold combinations are obtained during the stimulation process. .

[0172] The second baseline value acquisition module calculates the baseline value of the stimulation current intensity based on the user's assessment level, within a range defined by a personalized combination of current intensity thresholds, and incorporates a severity coefficient. .For example: when and At the time, at the threshold of the sensed current intensity and motion current intensity threshold Interpolation between: .

[0173] when and At the threshold of motion current intensity and the upper limit of withstand current intensity Interpolation between: .

[0174] when and At the threshold of motion current intensity and the upper limit of withstand current intensity Perform another type of interpolation: .

[0175] Based on the above, using the first baseline value acquisition module, and according to the above method (i.e., using the upper limit value, lower limit value, and the severity coefficient), other intervention stimulation parameters (including stimulation frequency) besides the stimulation current intensity are determined. Pulse width Ascent time Duration of stimulation Stop time Total stimulation duration The base value of (etc.).

[0176] The following is an example, assuming the user's personalized current intensity threshold combination is as follows: ; ; .when , The abnormal type is insufficient muscle recruitment, and according to The calculated severity coefficient within the interval is At that time, the combination of intervention stimulus parameters is as follows: ; ; ; ; ; ; .

[0177] when , The abnormality type is a complex functional decline, and At that time, the combination of intervention stimulus parameters is as follows: ; ; ; ; ; ; .

[0178] when , The abnormality type is insufficient local deformation of the larynx, and At that time, the combination of intervention stimulus parameters is as follows: ; ; ; ; ; ; .

[0179] This application embodiment separates the individual-specific current intensity parameter from the general parameter generation logic and anchors it to the user's own physiological feedback threshold, so that the main intervention stimulation parameter dynamically adapts to individual differences, ensuring that the stimulation current can achieve an effective therapeutic intensity without exceeding the user's comfort tolerance range.

[0180] In one possible implementation, the adjustment module includes a risk correction module configured to determine a risk correction coefficient based on the risk level, and to use the risk correction coefficient to correct the baseline values ​​of each of the intervention stimulus parameters to obtain the combination of intervention stimulus parameters.

[0181] Risk Adjustment Coefficient It can be represented as: (43); in, Let be the risk reduction coefficient, and satisfy: .

[0182] when When the risk is low, This means that the base value will not be modified, and the parameters will be output as originally planned.

[0183] when When this occurs, it indicates that there is currently a certain risk. At this point, the base value of the parameter will be multiplied by This allows for the reduction of stimulation intensity, such as reducing current intensity or shortening the total stimulation duration.

[0184] when When this occurs, it indicates that there is currently a high risk. This means directly correcting the baseline value of the parameter to 0, thereby forcibly stopping the intervention stimulus.

[0185] In one possible implementation, the final combination of intervention stimulus parameters can be determined by the following formula. : (44).

[0186] Similarly, for the above embodiment that determines the base value based on a personalized combination of current intensity thresholds, when At this time, the stimulation current intensity can be multiplied by the risk reduction factor. , and obtain the correction value.

[0187] The embodiments of this application transform discrete risk levels into continuous parameter adjustment coefficients, so that the response to risk is no longer a simple switching logic, but includes a multi-level safety strategy that includes normal output, derated output, and stop output, thereby improving the feasibility of the entire closed-loop intervention system.

[0188] In one possible implementation, the evaluation module 320 further includes an anomaly type determination module, which is configured to determine the anomaly type based on the electromyographic insufficiency component, the deformation insufficiency component, the time delay anomaly component, the coordination anomaly component, and the functional stability fluctuation risk component.

[0189] The adjustment module further includes a single-parameter adjustment module, which is configured to, after correcting the base values ​​of each of the intervention stimulus parameters using the risk correction coefficient, determine the target parameter category that needs to be adjusted individually in the combination of intervention stimulus parameters and the adjustment range of the intervention stimulus parameter of the target parameter category according to the anomaly type, and adjust the intervention stimulus parameter of the target parameter category according to the adjustment range to obtain the combination of intervention stimulus parameters.

[0190] Specifically, the anomaly type is part of the evaluation result output by the evaluation module 320, and the evaluation result can be expressed as follows: (45).

[0191] Exception types It can be determined based on the relationship between the various reduction components, and can be expressed as: (46); in, This indicates insufficient muscle recruitment. This indicates insufficient laryngeal deformation. This indicates a decline in complex functions. Indicates a timing coordination anomaly. This indicates a decrease in motion stability. , , , , These are the corresponding anomaly detection thresholds.

[0192] Specifically, the method for determining the exception type is as follows: when and At that time, it was determined that the current functional decline was mainly due to insufficient recruitment of swallowing-related muscle groups. ).in The threshold values ​​are the insufficient electromyographic component and the insufficient deformation component.

[0193] when and At that time, the current functional decline was mainly due to insufficient laryngeal elevation or insufficient local deformation. ).

[0194] when and At that time, it was determined that the current functional decline included a complex functional decline resulting from both insufficient muscle recruitment and insufficient local deformation of the larynx. ).

[0195] when and At that time, the current functional decline was mainly characterized by abnormal swallowing timing and decreased channel coordination. ).in The threshold values ​​are for time delay anomaly components and coordination anomaly components.

[0196] when At that time, it was determined that there was a decrease in the stability of the swallowing action. ).in This is the threshold for functional stability fluctuations.

[0197] The abnormality type determination module determines the abnormality type of this swallowing based on the above abnormality type determination criteria.

[0198] After receiving the abnormality type diagnosed above, the single-parameter adjustment module determines the target parameter category and specific adjustment range that need to be adjusted first and independently according to preset rules.

[0199] In one possible implementation, when the anomaly type is insufficient muscle recruitment, the current intensity is preferentially adjusted. or pulse width When the abnormality is due to insufficient local deformation of the larynx, the stimulation frequency should be adjusted preferentially. Duration of stimulation Or total stimulation duration When the abnormality type is a complex functional decline, the overall adjustment frequency is adjusted. Pulse width Current intensity Total stimulation duration When the abnormality is a swallowing timing disorder or decreased channel coordination, the rise time should be adjusted first. Stop time and duty cycle This makes the stimulus output rhythm smoother and more stable; when the abnormality type is signal quality abnormality or insufficient assessment confidence, the stimulus parameters are not increased, and a prompt for re-acquisition or manual review can be given.

[0200] It should be noted that the adjustment range in the single-parameter adjustment module can be manually entered or preset in the system.

[0201] Based on general safety adjustments, this application embodiment performs targeted adjustments based on the type of abnormality. It intelligently optimizes the internal structure of the intervention stimulus parameter combination according to the user's specific functional shortcomings, making the intervention stimulus not only safe but also more targeted.

[0202] In one possible implementation, the adjustment module further includes a critical adjustment module, which is configured to reduce each of the intervention stimulus parameters to obtain the combination of intervention stimulus parameters after adjusting the intervention stimulus parameters of the target parameter category according to the anomaly type, provided that the difference between the intervention risk score and the upper limit of the second target risk level is less than a difference threshold.

[0203] The second target risk level is higher than the first target risk level.

[0204] Specifically, the second target risk level corresponds to This means there is a certain risk involved.

[0205] During the continuous monitoring of the intervention risk score R, if at a certain moment R is still at the second target risk level ( It is within the range of ), but its value is already very close to the upper limit (i.e. If the difference between the two is less than a preset difference threshold, then critical adjustment is activated.

[0206] When critical regulation is activated, the critical regulation module reduces various intervention stimulus parameters. For example, it multiplies the already adjusted parameter values ​​by an additional reduction coefficient (preset value) to make a conservative parameter reduction before entering a higher risk level.

[0207] The embodiments of this application address the smoothing process near the critical point through adjustment, avoiding the output of high-intensity stimulation when approaching the risk zone, thereby enhancing safety redundancy and adaptability to dynamic changes in the user's physiological state.

[0208] Based on the above, this application also provides a method for monitoring and intervening in swallowing disorders. For example... Figure 6 As shown, methods for monitoring and intervening in dysphagia include: S610: Real-time reception of multimodal physiological signals collected by the multimodal sensing component during swallowing; S620: Obtain an assessment result characterizing the current swallowing function state based on the multimodal physiological signals; S630: Generate a combination of intervention stimulus parameters based on the evaluation results, wherein the combination of intervention stimulus parameters includes multiple intervention stimulus parameters; S640: Generate a stimulus control command based on the combination of intervention stimulus parameters, and send the stimulus control command to the stimulus component to drive the stimulus component to output intervention stimulus.

[0209] In practical use, the wearable base is first wrapped around the user's neck and secured via the connection area, aligning the throat sensing area with the user's front throat region. Then, the mandibular sensing area is attached to the wearable base via a detachable connection interface and fitted to the lower edge of the mandible. Next, hydrogel electrodes are installed at the corresponding electrode clip positions in the throat and / or mandibular sensing areas, ensuring stable contact between each sensing area and the user's skin.

[0210] During signal acquisition, a flexible strain sensor acquires local deformation signals of the larynx during swallowing, a triaxial accelerometer acquires vibration and acceleration signals of the larynx, and a hydrogel electrode acquires surface electromyographic signals of the larynx and / or mandibular region. The first microcontroller synchronously acquires, clock-aligns, preprocesses, and transmits the above multimodal signals to form multimodal physiological signals for swallowing function assessment.

[0211] After signal acquisition and preprocessing, based on surface electromyography (EMG) signals, local laryngeal deformation signals, and inertial change signals within the swallowing event window, EMG features, local laryngeal deformation features, inertial response features, swallowing timing features, and channel coordination features are extracted. Furthermore, a comprehensive score for swallowing function decline and an intervention risk score are constructed to determine the risk level and assessment level. Subsequently, an assessment result is generated based on the comprehensive swallowing function decline score, intervention risk score, risk level, and assessment level. Preferably, the abnormality type is also obtained and included as part of the assessment result.

[0212] When the assessment results show that the current swallowing function is basically stable, the intervention stimulation process is not initiated; only the assessment record is retained or monitoring continues. When the assessment results show mild, moderate, or significant functional decline, and the risk status allows for intervention, the intervention stimulation parameter combination is automatically determined based on the assessment results. The intervention stimulation parameters include at least stimulation frequency, pulse width, current intensity, rise time, stimulation duration, stop time, and total stimulation duration. Before using the stimulation intervention function for the first time, the system can also measure the user's personalized current intensity threshold combination and individualize the subsequent stimulation current intensity settings based on the sensory current intensity threshold, motor current intensity threshold, and the upper limit of the tolerated current intensity.

[0213] After determining the combination of intervention stimulation parameters, the intervention stimulation phase begins. Specifically, the host computer issues a control command, and the second microcontroller receives the control command, generates a corresponding stimulation control instruction, and controls the stimulation output pathway to output a biphasic pulse stimulation signal through the hydrogel electrode to provide auxiliary intervention to the swallowing-related target muscle groups. When the set total stimulation duration is reached or a stop command is received, the second microcontroller stops the stimulation output and records the comprehensive evaluation results, stimulation parameters, stimulation start time, and stimulation end time.

[0214] Therefore, this embodiment forms a complete working process of multimodal signal acquisition, swallowing function assessment, assessment result output, automatic setting of intervention stimulation parameters, and intervention stimulation output under the integrated monitoring and intervention structure, so that signal acquisition, function assessment and intervention control can be achieved collaboratively within the same wearable device framework.

[0215] In this embodiment, multimodal sensing and stimulation components are deployed in the larynx and mandibular regions. After evaluating the multimodal physiological signals using a processing unit, the evaluation results are converted into corresponding intervention stimulation parameters, and intervention stimulation is executed, forming a real-time closed-loop processing flow from signal acquisition to intervention stimulation. This not only comprehensively reflects the synergistic relationship between different physiological activities during swallowing through multimodal physiological signals, improving the accuracy of the evaluation, but also links swallowing function monitoring with rehabilitation intervention in real time, improving the precision and adaptability of the intervention.

[0216] Figure 7 This is a schematic diagram of the structure of the electronic device provided in this application, such as... Figure 7 As shown, the electronic device may include: a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, the communications interface 720, and the memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute a swallowing disorder monitoring and intervention control method, which includes: Real-time reception of multimodal physiological signals collected by multimodal sensing components during swallowing; An assessment result characterizing the current swallowing function status is obtained based on the multimodal physiological signals. Based on the evaluation results, a combination of intervention stimulus parameters is generated, which includes multiple intervention stimulus parameters. Stimulation control instructions are generated based on the combination of intervention stimulation parameters, and the stimulation control instructions are sent to the stimulation component to drive the stimulation component to output intervention stimulation.

[0217] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0218] On the other hand, this application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer is able to execute the swallowing disorder monitoring and intervention control method provided in the above embodiments. The method includes: Real-time reception of multimodal physiological signals collected by multimodal sensing components during swallowing; An assessment result characterizing the current swallowing function status is obtained based on the multimodal physiological signals. Based on the evaluation results, a combination of intervention stimulus parameters is generated, which includes multiple intervention stimulus parameters. Stimulation control instructions are generated based on the combination of intervention stimulation parameters, and the stimulation control instructions are sent to the stimulation component to drive the stimulation component to output intervention stimulation.

[0219] In another aspect, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the swallowing disorder monitoring and intervention control method provided in the above embodiments, the method comprising: Real-time reception of multimodal physiological signals collected by multimodal sensing components during swallowing; An assessment result characterizing the current swallowing function status is obtained based on the multimodal physiological signals. Based on the evaluation results, a combination of intervention stimulus parameters is generated, which includes multiple intervention stimulus parameters. Stimulation control instructions are generated based on the combination of intervention stimulation parameters, and the stimulation control instructions are sent to the stimulation component to drive the stimulation component to output intervention stimulation.

[0220] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0221] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0222] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A dysphagia monitoring and intervention device, characterized in that, Includes multimodal sensing components, stimulation components, and processing units; The multimodal sensing component is configured to acquire multimodal physiological signals during swallowing in real time; The stimulation component is configured to output interventional stimuli; The processing unit is electrically connected to the multimodal sensing component and the stimulation component, respectively; The processing unit includes a signal receiving module, an evaluation module, a parameter generation module, and a control module; The signal receiving module is configured to receive the multimodal physiological signals in real time; The assessment module is configured to obtain assessment results characterizing the current swallowing function state based on the multimodal physiological signals; The parameter generation module is configured to generate a combination of intervention stimulus parameters based on the evaluation results, and the combination of intervention stimulus parameters includes multiple intervention stimulus parameters. The control module is configured to control the stimulation component to output the intervention stimulus according to the combination of intervention stimulus parameters.

2. The dysphagia monitoring and intervention device of claim 1, wherein, The evaluation module includes a score determination module, a grade determination module, and an evaluation result acquisition module; The score determination module is configured to obtain a comprehensive score for swallowing dysfunction and an intervention risk score based on the multimodal physiological signals; The level determination module is configured to determine the risk level based on the intervention risk score, and to determine the assessment level based on the comprehensive score of the swallowing function decline and the intervention risk score; The assessment result acquisition module is configured to use a combination of the swallowing function impairment score, the intervention risk score, the risk level, and the assessment level as the assessment result.

3. The dysphagia monitoring and intervention device of claim 2, wherein, The parameter generation module includes a range determination module, a base value determination module, and an adjustment module; The range determination module is configured to determine the upper and lower limits of each of the intervention stimulus parameters corresponding to the assessment level when the risk level is the first target risk level; The baseline value determination module is configured to determine the baseline values ​​of each of the intervention stimulation parameters based on the comprehensive score of the swallowing function impairment. The adjustment module is configured to adjust the base value according to the risk level and the intervention risk score within the parameter range determined by the upper limit and the lower limit, so as to generate the intervention stimulus parameter combination.

4. The dysphagia monitoring and intervention device of claim 3, wherein, The multimodal physiological signals include local laryngeal deformation signals, surface electromyography signals, and inertial motion signals; The score determination module includes a feature extraction module, a swallowing function impairment component determination module, and a comprehensive swallowing function impairment score determination module; The feature extraction module is configured to extract electromyographic features from the surface electromyography signal; extract laryngeal local deformation features from the laryngeal local deformation signal; extract inertial response features from the inertial motion signal; and determine swallowing response delay features based on the peak time of any one of the laryngeal local deformation signal, surface electromyography signal, and inertial motion signal as a reference. Channel coordination features between the multimodal sensing components are extracted based on the multimodal physiological signals. The swallowing function impairment component determination module is configured to determine the electromyographic insufficiency component based on the electromyographic characteristics; determine the deformation insufficiency component based on the laryngeal local deformation characteristics; determine the inertial insufficiency component based on the inertial response characteristics; determine the time delay abnormal component based on the swallowing response time delay characteristics; and determine the coordination abnormal component based on the channel coordination characteristics. The swallowing function impairment comprehensive score determination module is configured to perform a weighted summation of the electromyographic insufficiency component, the deformation insufficiency component, the inertia insufficiency component, the time delay abnormality component, and the coordination abnormality component to obtain the swallowing function impairment comprehensive score.

5. The dysphagia monitoring and intervention device of claim 4, wherein, The score determination module also includes a risk component determination module and an intervention risk score determination module; The risk component determination module is configured to determine the risk components of abnormal outbreak signal after swallowing, abnormal recovery after swallowing, functional stability fluctuation risk components, and abnormal signal quality risk components based on the multimodal physiological signals. The intervention risk score is configured to be obtained by weighted summation of the risk components of the abnormal outbreak signal after swallowing, the abnormal recovery after swallowing, the functional stability fluctuation risk component, and the abnormal signal quality risk component.

6. The dysphagia monitoring and intervention device of claim 3, wherein, The baseline value determination module includes a severity coefficient determination module and a first baseline value acquisition module; The severity coefficient determination module is configured to determine the severity coefficient based on the comprehensive score of the swallowing function impairment; The first base value acquisition module is configured to obtain a base value for at least one of the intervention stimulus parameters based on the severity coefficient, the upper limit value, and the lower limit value of each of the intervention stimulus parameters.

7. The dysphagia monitoring and intervention device of claim 6, wherein, The combination of intervention stimulation parameters includes stimulation current intensity and at least one other intervention stimulation parameter; The baseline value determination module further includes a second baseline value acquisition module, which is configured to determine the baseline value of the stimulation current intensity based on the user's personalized current intensity threshold combination, assessment level, and severity coefficient. The personalized current intensity threshold combination includes a sensory current intensity threshold, a motor current intensity threshold, and an upper limit of tolerable current intensity. The sensory current intensity threshold represents the current intensity when the user first perceives the stimulus, the motor current intensity threshold represents the current intensity when the user begins to observe or detect a muscle response to the intervention stimulus, and the upper limit of tolerable current intensity represents the highest current intensity that the user can tolerate. Accordingly, the first base value acquisition module is configured to determine the base values ​​of the other intervention stimulus parameters based on the upper and lower limits of the other intervention stimulus parameters and the severity coefficient.

8. The dysphagia monitoring and intervention device of claim 5, wherein, The adjustment module includes a risk correction module, which is configured to determine a risk correction coefficient based on the risk level, and use the risk correction coefficient to correct the base values ​​of each of the intervention stimulus parameters to obtain the combination of intervention stimulus parameters.

9. The dysphagia monitoring and intervention device of claim 8, wherein, The assessment module further includes an anomaly type determination module, which is configured to determine the anomaly type based on the electromyographic insufficiency component, the deformation insufficiency component, the time delay anomaly component, the coordination anomaly component, and the functional stability fluctuation risk component. The adjustment module further includes a single-parameter adjustment module, which is configured to, after correcting the base values ​​of each of the intervention stimulus parameters using the risk correction coefficient, determine the target parameter category that needs to be adjusted individually in the combination of intervention stimulus parameters and the adjustment range of the intervention stimulus parameter of the target parameter category according to the anomaly type, and adjust the intervention stimulus parameter of the target parameter category according to the adjustment range to obtain the combination of intervention stimulus parameters.

10. The dysphagia monitoring and intervention device of claim 9, wherein, The adjustment module further includes a critical adjustment module, which is configured to reduce each of the intervention stimulus parameters after adjusting the intervention stimulus parameters of the target parameter category according to the abnormality type, when the difference between the intervention risk score and the upper limit of the second target risk level is less than a difference threshold, so as to obtain the combination of intervention stimulus parameters. The second target risk level is higher than the first target risk level.

11. The swallowing disorder monitoring and intervention device according to claim 5, characterized in that, The risk component determination module includes a first risk component determination module, which is configured to determine the electromyographic burst feature value and the inertial response energy within the target time window after swallowing based on the multimodal physiological signal, and to perform a weighted summation of the electromyographic burst feature value and the inertial response energy to obtain the risk component of the abnormal burst signal after swallowing.

12. The swallowing disorder monitoring and intervention device according to claim 5, characterized in that, The risk component determination module includes a second risk component determination module, which is configured to determine the first time required for the local laryngeal deformation signal to recover from its peak to its deformation baseline and the second time required for the inertial motion signal to recover from its peak to its inertial baseline, and to perform a weighted summation of the first time, the second time, and the time delay abnormal component to obtain the post-swallowing recovery abnormal risk component.

13. The swallowing disorder monitoring and intervention device according to claim 5, characterized in that, The risk component determination module includes a third risk component determination module, which is configured to determine the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature, and the swallowing response delay feature, respectively, based on the electromyographic feature, the laryngeal local deformation feature, the inertial response feature, and the swallowing response delay feature, and to perform a weighted summation of the electromyographic feature fluctuation component, the laryngeal local deformation feature fluctuation component, the inertial response feature fluctuation component, and the swallowing delay fluctuation component to obtain the functional stability fluctuation risk component.

14. The swallowing disorder monitoring and intervention device according to claim 5, characterized in that, The risk component determination module includes a fourth risk component determination module, which is configured to determine motion artifact risk components, channel loss risk components, and electrode contact stability risk components based on the multimodal physiological signals, and to perform a weighted summation of the motion artifact risk components, the channel loss risk components, and the electrode contact stability risk components to obtain the signal quality abnormality risk components.

15. A method for monitoring and intervening in dysphagia, characterized in that, Based on the swallowing disorder monitoring and intervention device according to any one of claims 1 to 14, the method comprises: Real-time reception of multimodal physiological signals collected by multimodal sensing components during swallowing; An assessment result characterizing the current swallowing function status is obtained based on the multimodal physiological signals. Based on the evaluation results, a combination of intervention stimulus parameters is generated, which includes multiple intervention stimulus parameters. Stimulation control instructions are generated based on the combination of intervention stimulation parameters, and the stimulation control instructions are sent to the stimulation component to drive the stimulation component to output intervention stimulation.

16. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the swallowing disorder monitoring and intervention control method as described in claim 15.

17. A non-transitory computer-readable storage medium, wherein a computer program is stored on the non-transitory computer-readable storage medium, characterized in that, When the computer program is executed by the processor, it implements the swallowing disorder monitoring and intervention control method as described in claim 15.

18. A computer program product, the computer program product comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the swallowing disorder monitoring and intervention control method as described in claim 15.

Images