Multi-dimensional swallowing rehabilitation training method and system
By combining visual, auditory, and gustatory stimulation in a multi-dimensional swallowing rehabilitation training method, and utilizing both EMG and swallowing acoustic signals for dual judgment, the problem of poor training effect and inaccurate assessment in existing technologies has been solved, achieving a more efficient swallowing rehabilitation training effect.
Patent Information
- Application Number
- CN202511619368.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-06
- Publication Date
- 2026-02-10
AI Technical Summary
Existing swallowing rehabilitation training methods suffer from poor training effects, low patient participation, and low accuracy in quantitative assessment of training effects, lacking multi-sensory stimulation and precise assessment methods.
Swallowing sounds are played through a speaker, liquid is injected through a micropump to induce swallowing, and EMG and swallowing acoustic signals are collected simultaneously for dual assessment. Combined with visual and gustatory stimulation, a multi-dimensional evaluation report is generated.
It increased the fun and initiative of training, achieved more accurate swallowing action assessment, and improved the effectiveness of rehabilitation training and patient compliance.
Smart Images
Figure CN121490337A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of rehabilitation medicine technology. More specifically, this invention relates to a multi-dimensional swallowing rehabilitation training method and system. Background Technology
[0002] Dysphagia is a serious complication that can result from various diseases, affecting patients' normal eating and quality of life. Swallowing rehabilitation training is an important medical approach to help patients restore their swallowing function. Its core lies in reshaping the patient's swallowing neural pathways and muscle coordination through repetitive and standardized movement training.
[0003] However, existing swallowing rehabilitation training methods have several technical problems. First, traditional training methods often rely on video imitation or simple verbal instructions, lacking multi-dimensional sensory stimulation such as taste and hearing. This results in a serious disconnect between the training scenario and the actual eating process, leading to unsatisfactory training effects and patient participation. Second, existing technologies generally lack objective and precise quantitative methods for evaluating training effectiveness. To address this issue, the industry has begun to explore the introduction of physiological signal acquisition and analysis technologies. For example, surface electromyography (sEMG) signals are collected to determine the effectiveness of swallowing movements. However, relying solely on sEMG signals for assessment is highly susceptible to interference from environmental noise, muscle fatigue, or individual vocal differences, resulting in insufficient accuracy and stability of the evaluation results.
[0004] In summary, existing swallowing rehabilitation training methods suffer from technical problems such as poor training effectiveness, low patient participation, and poor accuracy in quantitative assessment of training effectiveness.
[0005] Therefore, how to construct a multi-sensory, immersive training scenario and develop a more robust and accurate method for quantifying swallowing movements are technical challenges that urgently need to be addressed in this field. Summary of the Invention
[0006] To address the technical problems of poor training effectiveness, low patient participation, and inaccurate quantitative assessment of training results in existing swallowing rehabilitation training methods, this invention provides solutions in the following aspects.
[0007] In a first aspect, the present invention provides a multi-dimensional swallowing rehabilitation training method, comprising: The standard swallowing action is demonstrated to the patient through a speaker, swallowing sound effects synchronized with the swallowing action are played, and liquid is injected into the patient's mouth using a micropump and catheter at the beginning of the standard swallowing action to induce the patient to swallow naturally. EMG signals of the patient's swallowing muscles and swallowing acoustic signals were collected in real time during the patient's swallowing action. The first judgment result is obtained by determining whether the current swallowing is effective based on the EMG signal of the swallowing muscle group; the second judgment result is obtained by determining whether the current swallowing is effective based on the swallowing acoustic signal. Specifically, determining whether the current swallowing is effective based on the EMG signal of the swallowing muscle group includes: if the amplitude of the EMG signal exceeds a threshold, the first judgment result is effective swallowing; otherwise, the first judgment result is invalid swallowing. Determining whether the current swallowing is effective based on the swallowing acoustic signal includes: if the duration of the acquired swallowing acoustic signal is greater than 0.5 seconds and the total energy of the swallowing acoustic signal is greater than a total energy threshold and the center frequency of the swallowing acoustic signal is less than a center frequency threshold, the first judgment result is effective swallowing; otherwise, the first judgment result is invalid swallowing. If both the first and second judgment results are valid swallowing, then the current swallowing is determined to be valid swallowing; otherwise, the current swallowing is determined to be invalid swallowing. A successful swallow indicates that the patient has swallowed successfully; an ineffective swallow indicates that the patient has failed to swallow. In response to the patient reaching the total number of swallowing training sessions for the current group, an assessment report is generated and displayed. The assessment report includes swallowing frequency, which is used to characterize the frequency with which the patient performs effective swallowing during the current group of swallowing training sessions.
[0008] Preferably, the expression for calculating the swallowing frequency is: ; In the formula, f represents the swallowing frequency, n represents the number of effective swallows by the patient in the current group of swallowing training, and T represents the time taken for the patient to reach the total number of swallows in the current group of swallowing training.
[0009] Preferably, it further includes: Calculate the patient's completion rate of the current swallowing training. The completion rate of the current swallowing training is equal to the number of effective swallows by the patient in the current swallowing training divided by the total number of swallows in the current swallowing training. If the completion rate of swallowing training is less than 50%, increase the volume of liquid injected into the patient's mouth at the beginning of the standard swallowing action when performing the next set of swallowing training.
[0010] Preferably, the assessment report also includes the swallowing completion rate for each swallow, and the swallowing completion rate for the current swallow. The calculation expression is: ; In the formula, This indicates the displacement of the Adam's apple during the current swallow. This indicates the displacement of the Adam's apple during effective swallowing in healthy individuals.
[0011] Preferably, the method for obtaining the laryngeal elevation displacement value during the previous swallow includes: A three-axis gyroscope fixed to the patient's larynx was used to collect the angular velocity of the larynx along the X-axis at various moments during the entire swallowing process of the current swallow; the X-axis is an axis that extends in the left and right direction and passes through the patient's throat. Calculate the maximum angle rotated vertically along the line connecting the Adam's apple and the center of rotation during the entire swallowing process based on the angular velocity of the Adam's apple along the X-axis at each moment. The center of rotation is located on the X-axis; Using the length of the line connecting the three-axis gyroscope and the center of rotation as the radius of rotation, and calculating the displacement of the Adam's apple during the current swallow, the calculation expression is as follows: ; In the formula, ΔY represents the displacement of the Adam's apple during the current swallow. Indicates the radius of rotation. It represents the maximum angle that the line connecting the Adam's apple and the center of rotation rotates through in the vertical direction during the entire swallowing process.
[0012] Preferably, the assessment report also includes the delay time for each swallow, which is the time interval between the injection of liquid into the patient's mouth and the detection of larynx elevation.
[0013] Preferably, the evaluation report also includes an evaluation score S for each swallow. The higher the evaluation score S, the more successful the swallowing action. The calculation expression is as follows: ; In the formula, , and All are weighting coefficients. This represents the ratio of the RMS value of the EMG signal during the current swallow to the RMS value of the EMG signal during the maximum voluntary contraction. CI indicates the similarity between the swallowing acoustic signal of the current swallow and the swallowing acoustic signal of the standard swallowing action, while CI indicates the swallowing completion rate of the current swallow.
[0014] Preferably, the method for obtaining the similarity between the swallowing acoustic signal of the current swallow and the swallowing acoustic signal of a standard swallowing action includes: Preprocessing of the swallowing acoustic signal of the current swallow includes: noise reduction and framing; Feature extraction was performed on the preprocessed swallowing acoustic signal to obtain the first MFCC feature sequence; The cumulative distance is obtained by comparing the first MFCC feature sequence with the MFCC feature sequence of the swallowing acoustic signal of the standard swallowing action using the DTW algorithm. The cumulative distance is normalized to obtain the similarity.
[0015] Preferably, the evaluation report also includes the liquid swallowing flow rate for each swallow, and the method for obtaining the liquid swallowing flow rate for a given swallow includes: A matrix pressure sensor is used to collect the pressure of the liquid pushed backward by the base of the tongue in real time, while a three-axis gyroscope is used to collect the angular velocity of the Adam's apple rotating along the X-axis at various moments; the data acquisition frequency of the three-axis gyroscope is the same as that of the matrix pressure sensor and the start time of the acquisition is the same. The onset of swallowing is determined based on the collected pressure. The angle between the first and second straight lines at each moment after the initial moment is calculated based on the angular velocity of the Adam's apple rotating along the X-axis. The first straight line is the line connecting the Adam's apple to the center of rotation at that moment, and the second straight line is the line connecting the initial position of the Adam's apple to the center of rotation, thereby determining the termination moment of swallowing. The duration from the start of swallowing to the end of swallowing is taken as the duration of liquid passage through the pharynx; Calculating the swallowing flow rate includes: using the quotient of the volume of swallowed liquid and the time the liquid passes through the pharynx as the liquid swallowing flow rate for the current swallow.
[0016] In a second aspect, the present invention provides a multi-dimensional swallowing rehabilitation training system, comprising: A liquid injection device is used to inject liquid into the patient's mouth at the onset of a standard swallowing action to induce the patient's natural swallowing; the liquid injection device includes a micropump and a catheter; A speaker is used to play swallowing sounds synchronized with the swallowing action; A display screen is used to show patients standard swallowing movements; Electromyography (EMG) devices are used to collect EMG signals from the patient's swallowing muscles in real time. A microphone is used to collect the acoustic signals of swallowing during the patient's swallowing process; The host computer is connected to an electromyography (EMG) device and a microphone, and controls a speaker, a display screen, and a micropump to execute the multidimensional swallowing rehabilitation training method of the present invention.
[0017] The beneficial effects of this invention are as follows: Compared with existing training methods that rely solely on video, sound, or electrical stimulation for single-dimensional guidance, this invention combines visual and auditory guidance with gustatory stimulation (liquid injection) to actively induce a swallowing reflex that is closer to the patient's actual swallowing reflex, lowering the threshold of the swallowing reflex and increasing the initiative and enjoyment of training. At the same time, by combining surface electromyography signals and swallowing acoustic signals for dual judgment, the effectiveness of each swallowing action can be assessed more accurately and objectively, overcoming the shortcomings of traditional methods that lack intelligent assessment and cannot accurately judge the quality of training, thereby improving the overall effect of rehabilitation training and patient compliance. Attached Figure Description
[0018] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a schematic flowchart illustrating a multi-dimensional swallowing rehabilitation training method according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the structure of a multi-dimensional swallowing rehabilitation training system according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the structure of a host computer according to an embodiment of the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0021] Examples of multidimensional swallowing rehabilitation training methods: like Figure 1 As shown, the multidimensional swallowing rehabilitation training method of the present invention includes: S101. Inducing the patient's natural swallowing, specifically: displaying the standard swallowing action to the patient through a monitor, playing swallowing sound effects synchronized with the swallowing action, and injecting liquid into the patient's mouth using a micropump and catheter at the beginning of the standard swallowing action to induce the patient's natural swallowing. The liquid injected into the patient's mouth is a harmless, swallowable liquid, such as milk or other beverages. The taste of the swallowable liquid can be sour, sweet, salty, bitter, or similar.
[0022] Standard swallowing actions include drinking water, chewing, and swallowing. The swallowing sound effect can be a "gulp" sound.
[0023] A reservoir can be used to store the fluid to be injected into the patient's mouth. The input end of the micropump is sealed to the reservoir, and the output end is sealed to the tip of the catheter. The catheter tip is then safely and comfortably placed in the patient's mouth. The catheter tip is typically placed on the back of the patient's tongue or the front of the hard palate. This position best triggers oral sensory stimulation and the swallowing reflex. The catheter needs to be properly secured (e.g., with a support or by the patient gently biting down) to ensure it does not move or injure the patient during training.
[0024] S102. During the patient's swallowing action, EMG signals of the patient's swallowing muscle group and swallowing acoustic signals are collected in real time. An electromyography (EMG) device can be used to collect EMG signals from the patient's swallowing muscles. A microphone can be used to collect the patient's swallowing acoustic signals.
[0025] S103. Obtain the first judgment result and the second judgment result, specifically: determine whether the current swallowing is effective based on the EMG signal of the swallowing muscle group, thereby obtaining the first judgment result; determine whether the current swallowing is effective based on the swallowing acoustic signal, thereby obtaining the second judgment result; wherein, determining whether the current swallowing is effective based on the EMG signal of the swallowing muscle group includes: if the amplitude of the electromyographic signal exceeds the threshold, the first judgment result is effective swallowing, otherwise the first judgment result is invalid swallowing; determining whether the current swallowing is effective based on the swallowing acoustic signal includes: if the duration of the acquired swallowing acoustic signal is greater than 0.5 seconds and the total energy of the swallowing acoustic signal is greater than the total energy threshold and the center frequency of the swallowing acoustic signal is less than the center frequency threshold, the first judgment result is effective swallowing, otherwise the first judgment result is invalid swallowing; The sound waves and vibration signals produced by effective swallowing and sham swallowing differ fundamentally in their physical characteristics. The acoustic signal components of effective swallowing have lower frequencies and higher energy, sounding more "moist" and "dull." In contrast, the acoustic signal components of sham swallowing have higher frequencies and lower energy, sounding more "dry" and "crisp." Furthermore, the duration of the acoustic signal from effective swallowing is typically longer than that from sham swallowing, the signal amplitude is usually higher, and the area under the acoustic signal curve (i.e., the total energy of the signal) is larger than that of sham swallowing. Moreover, compared to sham swallowing, the center frequency of the acoustic signal from effective swallowing is lower; the center frequency refers to the frequency point where the signal energy is most concentrated.
[0026] The center frequency threshold can be determined experimentally.
[0027] When a patient performs a swallowing action, the contraction strength of their swallowing muscles is relatively large (corresponding to a larger amplitude of electromyography signal); otherwise, the contraction strength of their swallowing muscles is relatively small (corresponding to a smaller amplitude of electromyography signal). Therefore, by comparing the amplitude of the electromyography signal with the threshold, it is possible to determine more accurately whether the patient's previous swallowing was effective.
[0028] S104. Determine whether the current swallow is a valid swallow, specifically: if both the first and second determination results are valid swallows, then determine that the current swallow is a valid swallow; otherwise, determine that the current swallow is an invalid swallow. Since the EMG signal acquisition device and the swallowing acoustic signal acquisition device for the patient's swallowing muscles may be affected by noise during the acquisition process, resulting in errors in the acquired data, the determination of swallowing effectiveness can be further improved by combining the first and second judgment results to determine whether swallowing is effective.
[0029] S105. If the current swallow is a successful swallow, it indicates that the patient has swallowed successfully; if the current swallow is an ineffective swallow, it indicates that the patient has failed to swallow. Providing prompts to patients can help them correct their swallowing movements the next time they swallow.
[0030] S106. Generate and display an assessment report, specifically: in response to the patient's swallowing training reaching the total number of swallows in the current group, generate and display an assessment report. The assessment report includes swallowing frequency, which is used to characterize the frequency with which the patient performs effective swallowing in the current group of swallowing training.
[0031] Generating an assessment report after the current group's swallowing training is completed helps patients and doctors to intuitively understand the patient's swallowing ability recovery.
[0032] This method injects liquid into the patient's mouth, thereby using taste to strongly activate the brain's swallowing center, lowering the "threshold" that triggers the swallowing reflex, making swallowing easier and faster. Combining taste and auditory feedback simulates a real swallowing scenario, allowing for better training of swallowing ability. By playing swallowing instruction videos and sound effects to guide the patient's swallowing movements, real-time visual and auditory feedback is provided, engaging multiple senses during the swallowing process. This allows for comprehensive assessment of swallowing ability and more accurate judgment of the quality of the swallowing action.
[0033] Compared to existing training methods that rely solely on video, sound, or electrical stimulation for single-dimensional guidance, this invention combines visual and auditory guidance with gustatory stimulation (liquid injection) to actively induce a more realistic swallowing reflex in patients, lowering the swallowing reflex threshold and increasing the initiative and enjoyment of training. Simultaneously, by combining surface electromyography signals and swallowing acoustic signals for dual assessment, the effectiveness of each swallowing action can be evaluated more accurately and objectively, overcoming the shortcomings of traditional methods that lack intelligent assessment and cannot accurately judge training quality. This improves the overall effectiveness of rehabilitation training and patient compliance.
[0034] In this embodiment, the expression for calculating the swallowing frequency is: ; In the formula, f represents the swallowing frequency, n represents the number of effective swallows by the patient in the current group of swallowing training, and T represents the time taken for the patient to reach the total number of swallows in the current group of swallowing training.
[0035] In one embodiment, it also includes: S201. Calculate the patient's completion rate of the current group of swallowing training. The completion rate of the current group of swallowing training is equal to the number of effective swallows by the patient in the current group of swallowing training divided by the total number of swallows in the current group of swallowing training. S202. In response to a completion rate of less than 50% in swallowing training, increase the volume of liquid injected into the patient's mouth at the beginning of the standard swallowing action when performing the next set of swallowing training.
[0036] Existing rehabilitation training typically has a fixed, preset difficulty level. This invention, however, can automatically increase the volume of fluid injected in the next training session based on the patient's real-time performance (completion rate below 50%), thus dynamically adjusting the training difficulty. This personalized, adaptive training method continuously provides patients with appropriate challenges, avoiding poor training results due to excessively high or low difficulty, thereby improving the efficiency of rehabilitation training and patient participation. In one embodiment, the assessment report also includes the swallowing completion rate for each swallow, and the swallowing completion rate for the previous swallow. The calculation expression is: ; In the formula, This indicates the displacement of the Adam's apple during the current swallow. This represents the upward displacement of the larynx during effective swallowing in healthy individuals. The upward displacement of the larynx during effective swallowing in healthy individuals can be set to 2.0 cm.
[0037] Compared to existing technologies that rely on therapists' visual observation or cannot be quantified in real time, this invention innovatively uses a three-axis gyroscope to accurately capture and calculate the key displacement value of larynx elevation. This method provides an objective and quantifiable swallowing completion index (CI) that accurately reflects the core biomechanical processes of swallowing, making the assessment results more convincing and scientific, and solving the problem that traditional methods are difficult to quantify and analyze key swallowing movements.
[0038] In this embodiment, the method for obtaining the laryngeal elevation displacement value during the previous swallow includes: S301. Use a three-axis gyroscope fixed to the patient's larynx to collect the angular velocity of the larynx along the X-axis at various moments during the entire swallowing process of the current swallow; the X-axis is an axis that extends in the left and right direction and passes through the patient's throat. The left and right directions refer to the left and right directions from the patient's perspective.
[0039] A three-axis gyroscope has three axes: the X-axis, the Y-axis, and the Z-axis. The X-axis (coronal axis) extends laterally and passes through the patient's throat; during swallowing, it corresponds to the forward and backward tilt of the Adam's apple. The Y-axis (vertical axis) extends vertically; rotation around this axis corresponds to the head's rotation, but its effect during swallowing is relatively small. The Z-axis (sagittal axis) extends anteriorly and posteriorly; rotation around this axis corresponds to the head's left and right tilt, but its effect during swallowing is also relatively small.
[0040] The primary movement associated with Adam's apple elevation is rotation around the X-axis. When the Adam's apple rises, it exhibits a forward and upward "tilting" or "rotating" motion. During swallowing, the gyroscope continuously outputs angular velocity values on three axes. , and The angular velocity around the X-axis when the Adam's apple begins to rise. There will be a noticeable pulse signal.
[0041] S302. Calculate the maximum angle rotated vertically by the line connecting the Adam's apple and the center of rotation during the entire swallowing process, based on the angular velocity of the Adam's apple along the X-axis at each moment. The center of rotation is located on the X-axis; In this embodiment, a point on the hyoid bone or cervical vertebrae can be selected as the center of rotation.
[0042] In this embodiment, the maximum angle The calculation expression is: ; In the formula, and These represent the start and end times of the current swallow, respectively. This represents the angular velocity of the Adam's apple along the X-axis at time t. Indicates in The angular velocity of the Adam's apple in the X-axis direction at any given moment. This indicates the duration of the angular velocity sampling interval (i.e., the time interval between two adjacent sampling points).
[0043] S303. Take the length of the line connecting the three-axis gyroscope and the center of rotation as the radius of rotation, and calculate the displacement of the Adam's apple during the current swallow. The calculation expression is: ; In the formula, ΔY represents the displacement of the Adam's apple during the current swallow. Indicates the radius of rotation. It represents the maximum angle that the line connecting the Adam's apple and the center of rotation rotates through in the vertical direction during the entire swallowing process.
[0044] Compared to traditional video fluorescence swallowing examination (VFSS), which involves radiation and high equipment requirements, or manual palpation by therapists, which suffers from subjectivity and low accuracy, using a gyroscope is a non-invasive, portable measurement method that can output objective data in real time. It can conveniently and accurately capture the movement angle of the larynx, thereby calculating the upward displacement, providing a reliable technical means for quantitative assessment.
[0045] In one embodiment, the assessment report also includes the delay time for each swallow, which is the time interval between the moment the liquid is injected into the patient's mouth and the detection of the patient's larynx rising.
[0046] For a valid swallowing action, the delay time is usually less than 1 second; for an invalid swallowing action, the delay time is usually more than 1 second.
[0047] The swallowing delay time, a parameter that measures the reaction time from stimulation (liquid injection) to swallowing initiation (larynx elevation), is an important indicator for clinically assessing aspiration risk. Compared to existing technologies that only focus on the swallowing action itself, this invention, by quantifying this time interval, makes rehabilitation assessment more comprehensive, effectively monitors the recovery of the patient's swallowing reflex arc, and provides objective evidence for aspiration prevention.
[0048] In one embodiment, the assessment report also includes an evaluation score S for each swallow. A higher evaluation score S indicates a more successful swallowing action, and its calculation expression is as follows: ; In the formula, , and All are weighting coefficients. This represents the ratio of the RMS value of the EMG signal during the current swallow to the RMS value of the EMG signal during the maximum voluntary contraction. CI indicates the similarity between the swallowing acoustic signal of the current swallow and the swallowing acoustic signal of the standard swallowing action, while CI indicates the swallowing completion rate of the current swallow.
[0049] Compared to existing technologies that analyze various physiological signals in isolation, this invention generates a comprehensive evaluation score (S) by weighted summation of electromyographic signal intensity, laryngeal elevation completion, and acoustic signal similarity. This comprehensive score can more comprehensively and three-dimensionally reflect the overall quality of a single swallowing action, avoiding the one-sidedness of single-indicator assessment, and providing therapists and patients with a more intuitive and comprehensive reference for rehabilitation progress.
[0050] Methods for obtaining the RMS value of the EMG signal at maximum spontaneous contraction include: S601. The patient performs a maximum isometric contraction of the muscle to be tested and the EMG signal during this maximum voluntary contraction is collected.
[0051] It is essential to ensure that the target muscle exerts its maximum voluntary effort and to provide ample verbal encouragement. For data reliability, it is typically necessary to repeat the exercise three times, with a two-minute rest between each repetition, and use the maximum value as the baseline.
[0052] S602. Process the EMG signal of maximum spontaneous contraction to obtain the baseline intensity value, including: Signal processing: Perform standard processing (filtering, full-wave rectification) on the acquired raw MVC signal.
[0053] Calculate the strength: Calculate the root mean square (RMS) value during the most stable phase of MVC contraction (usually a middle segment, such as a 500-millisecond window).
[0054] The method for obtaining the RMS value of the EMG signal from the previous swallow includes: filtering and rectifying the EMG signal from the previous swallow, and calculating its RMS value during the main contraction phase.
[0055] In one embodiment, the method for obtaining the similarity between the swallowing acoustic signal of the previous swallow and the swallowing acoustic signal of a standard swallowing action includes: S401. Preprocess the swallowing acoustic signal of the current swallow, including noise reduction and framing. S402. Extract features from the preprocessed swallowing acoustic signal to obtain the first MFCC feature sequence; S403. The first MFCC feature sequence is compared with the MFCC feature sequence of the swallowing acoustic signal of the standard swallowing action using the DTW algorithm to obtain the cumulative distance. S404. Normalize the cumulative distance to obtain the similarity. The corresponding calculation expression is: ; In the formula, Indicates the similarity, Indicates the cumulative distance. This represents the maximum possible distance that is preset; this value is usually determined experimentally and can be the distance obtained from the worst match (e.g., a typical swallowing signal and a non-swallowing noise signal).
[0056] The method in this embodiment calculates the similarity of swallowing acoustic signals using MFCC feature extraction and DTW algorithm. Compared to traditional methods that only compare simple acoustic features such as energy and duration, MFCC and DTW are mature algorithms in the field of speech recognition. They can more accurately capture and compare the time-domain and frequency-domain features of swallowing sounds, thereby more robustly and accurately judging the difference between the patient's swallowing acoustic pattern and the standard pattern, improving the scientific nature and accuracy of acoustic analysis.
[0057] In one embodiment, the assessment report also includes the liquid swallowing flow rate for each swallow, and the method for obtaining the liquid swallowing flow rate for a given swallow includes: S501: A matrix pressure sensor is used to collect the pressure of the liquid pushed backward by the base of the tongue in real time, and a three-axis gyroscope is used to collect the angular velocity of the Adam's apple rotating along the X-axis at various times; the data acquisition frequency of the three-axis gyroscope is the same as the data acquisition frequency of the matrix pressure sensor and the start time of acquisition is the same. S502. Determine the start time of swallowing based on the collected pressure; When the tongue begins to forcefully push the liquid backward, the pressure signal experiences a sharp and significant increase. The starting point of this increase is defined as the initiation point of swallowing.
[0058] S503. Calculate the angle between the first straight line and the second straight line at each moment based on the angular velocity of the Adam's apple rotating along the X-axis after the initial moment. The first straight line is the line connecting the Adam's apple and the center of rotation at that moment, and the second straight line is the line connecting the initial position of the Adam's apple and the center of rotation, thereby determining the termination moment of swallowing. During swallowing, the Adam's apple first rises to its highest point and then begins to descend. The moment when the Adam's apple returns to its resting baseline position is defined as the end point of swallowing. This point signifies the completion of the swallowing action, and the pharyngeal structures return to their original positions.
[0059] S504. The duration from the start of swallowing to the end of swallowing is taken as the duration of liquid passage through the pharynx. S505. Calculate the swallowing flow rate, including: taking the quotient of the volume of swallowed liquid and the duration of liquid passage in the pharynx as the liquid swallowing flow rate for the current swallow.
[0060] By combining data from pressure sensors and gyroscopes to determine the duration of liquid transit in the pharynx and calculate the flow rate, this embodiment assesses swallowing function from the perspective of bolus transport efficiency. This adds a completely new dimension to rehabilitation assessment, going beyond the analysis of muscle and skeletal movements to directly quantify the ultimate goal of swallowing—efficient liquid transport—making the assessment more comprehensive.
[0061] In one embodiment, the evaluation report also includes a comprehensive score for the current group training, which includes the mean swallowing completion rate for each swallow, the mean delay time for each swallow, and the mean evaluation score for each swallow.
[0062] Example of a multidimensional swallowing rehabilitation training system: This invention also provides a multi-dimensional swallowing rehabilitation training system. For example... Figure 2 As shown, the multidimensional swallowing rehabilitation training system includes: A liquid injection device is used to inject liquid into the patient's mouth at the onset of a standard swallowing action to induce the patient's natural swallowing; the liquid injection device includes a micropump and a catheter; A speaker is used to play swallowing sounds synchronized with the swallowing action; A display screen is used to show patients standard swallowing movements; Electromyography (EMG) devices are used to collect EMG signals from the patient's swallowing muscles in real time. A microphone is used to collect the acoustic signals of swallowing during the patient's swallowing process; The host computer is connected to the electromyography device and microphone, and the control is connected to the speaker, display screen and micropump to execute the multi-dimensional swallowing rehabilitation training method described in the above embodiments.
[0063] like Figure 3 As shown, the host computer includes a processor and a memory. The memory stores computer program instructions. When the computer program instructions are executed by the processor, the multi-dimensional swallowing rehabilitation training method described in the above embodiments is implemented.
[0064] The host computer also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.
[0065] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.
Claims
1. A multidimensional swallowing rehabilitation training method, characterized in that, include: The standard swallowing action is displayed to the patient on the screen, swallowing sound effects are played in sync with the swallowing action, and liquid is injected into the patient's mouth using a micropump and catheter at the beginning of the standard swallowing action to induce the patient to swallow naturally. EMG signals of the patient's swallowing muscles and swallowing acoustic signals were collected in real time during the patient's swallowing action. The first judgment result is obtained by determining whether the current swallowing is effective based on the EMG signal of the swallowing muscle group; the second judgment result is obtained by determining whether the current swallowing is effective based on the swallowing acoustic signal. Specifically, determining whether the current swallowing is effective based on the EMG signal of the swallowing muscle group includes: if the amplitude of the EMG signal exceeds a threshold, the first judgment result is effective swallowing; otherwise, the first judgment result is invalid swallowing. Determining whether the current swallowing is effective based on the swallowing acoustic signal includes: if the duration of the acquired swallowing acoustic signal is greater than 0.5 seconds and the total energy of the swallowing acoustic signal is greater than a total energy threshold and the center frequency of the swallowing acoustic signal is less than a center frequency threshold, the first judgment result is effective swallowing; otherwise, the first judgment result is invalid swallowing. If both the first and second judgment results are valid swallowing, then the current swallowing is determined to be valid swallowing; otherwise, the current swallowing is determined to be invalid swallowing. A successful swallow indicates that the patient has swallowed successfully; an ineffective swallow indicates that the patient has failed to swallow. In response to the patient reaching the total number of swallowing training sessions for the current group, an assessment report is generated and displayed. The assessment report includes swallowing frequency, which is used to characterize the frequency with which the patient performs effective swallowing during the current group of swallowing training sessions.
2. The multidimensional swallowing rehabilitation training method as described in claim 1, characterized in that, The formula for calculating the swallowing frequency is: ； In the formula, f represents the swallowing frequency, n represents the number of effective swallows by the patient in the current group of swallowing training, and T represents the time taken for the patient to reach the total number of swallows in the current group of swallowing training.
3. The multidimensional swallowing rehabilitation training method as described in claim 1, characterized in that, Also includes: Calculate the patient's completion rate of the current swallowing training. The completion rate of the current swallowing training is equal to the number of effective swallows by the patient in the current swallowing training divided by the total number of swallows in the current swallowing training. If the completion rate of swallowing training is less than 50%, increase the volume of liquid injected into the patient's mouth at the beginning of the standard swallowing action when performing the next set of swallowing training.
4. The multidimensional swallowing rehabilitation training method as described in claim 1, characterized in that, The assessment report also includes the swallowing completion rate for each swallow, and the swallowing completion rate for the previous swallow. The calculation expression is: ； In the formula, This indicates the displacement of the Adam's apple during the current swallow. This indicates the displacement of the Adam's apple during effective swallowing in healthy individuals.
5. The multidimensional swallowing rehabilitation training method as described in claim 4, characterized in that, Methods for obtaining the displacement value of the Adam's apple during the previous swallow include: A three-axis gyroscope fixed to the patient's larynx was used to collect the angular velocity of the larynx along the X-axis at various moments during the entire swallowing process of the current swallow; the X-axis is an axis that extends in the left and right direction and passes through the patient's throat. Calculate the maximum angle rotated vertically along the line connecting the Adam's apple and the center of rotation during the entire swallowing process based on the angular velocity of the Adam's apple along the X-axis at each moment. The center of rotation is located on the X-axis; Using the length of the line connecting the three-axis gyroscope and the center of rotation as the radius of rotation, and calculating the displacement of the Adam's apple during the current swallow, the calculation expression is as follows: ； In the formula, ΔY represents the displacement of the Adam's apple during the current swallow. Indicates the radius of rotation. It represents the maximum angle that the line connecting the Adam's apple and the center of rotation rotates through in the vertical direction during the entire swallowing process.
6. The multidimensional swallowing rehabilitation training method as described in claim 4, characterized in that, The assessment report also includes the delay time for each swallow, which is the time interval between the injection of liquid into the patient's mouth and the detection of larynx elevation.
7. The multidimensional swallowing rehabilitation training method as described in claim 6, characterized in that, The assessment report also includes a score S for each swallow. A higher score S indicates a more successful swallowing action. The calculation formula is as follows: ； In the formula, , and All are weighting coefficients. This represents the ratio of the RMS value of the EMG signal during the current swallow to the RMS value of the EMG signal during the maximum voluntary contraction. CI indicates the similarity between the swallowing acoustic signal of the current swallow and the swallowing acoustic signal of the standard swallowing action, while CI indicates the swallowing completion rate of the current swallow.
8. The multidimensional swallowing rehabilitation training method as described in claim 7, characterized in that, Methods for obtaining the similarity between the swallowing acoustic signal of the current swallow and the swallowing acoustic signal of the standard swallowing action include: Preprocessing of the swallowing acoustic signal of the current swallow includes: noise reduction and framing; Feature extraction was performed on the preprocessed swallowing acoustic signal to obtain the first MFCC feature sequence; The cumulative distance is obtained by comparing the first MFCC feature sequence with the MFCC feature sequence of the swallowing acoustic signal of the standard swallowing action using the DTW algorithm. The cumulative distance is normalized to obtain the similarity.
9. The multidimensional swallowing rehabilitation training method as described in any one of claims 1 to 8, characterized in that, The assessment report also includes the liquid swallowing flow rate for each swallow, and the methods for obtaining the liquid swallowing flow rate for a given swallow include: A matrix pressure sensor is used to collect the pressure of the liquid pushed backward by the base of the tongue in real time, while a three-axis gyroscope is used to collect the angular velocity of the Adam's apple rotating along the X-axis at various moments; the data acquisition frequency of the three-axis gyroscope is the same as that of the matrix pressure sensor and the start time of the acquisition is the same. The onset of swallowing is determined based on the collected pressure. The angle between the first and second straight lines at each moment after the initial moment is calculated based on the angular velocity of the Adam's apple rotating along the X-axis. The first straight line is the line connecting the Adam's apple to the center of rotation at that moment, and the second straight line is the line connecting the initial position of the Adam's apple to the center of rotation, thereby determining the termination moment of swallowing. The duration from the start of swallowing to the end of swallowing is taken as the duration of liquid passage through the pharynx; Calculating the swallowing flow rate includes: using the quotient of the volume of swallowed liquid and the time the liquid passes through the pharynx as the liquid swallowing flow rate for the current swallow.
10. A multi-dimensional swallowing rehabilitation training system, characterized in that, include: A liquid injection device is used to inject liquid into the patient's mouth at the onset of a standard swallowing action to induce the patient's natural swallowing; the liquid injection device includes a micropump and a catheter; A speaker is used to play swallowing sounds synchronized with the swallowing action; A display screen is used to show patients standard swallowing movements; Electromyography (EMG) devices are used to collect EMG signals from the patient's swallowing muscles in real time. A microphone is used to collect the acoustic signals of swallowing during the patient's swallowing process; The host computer is connected to the electromyography device and microphone, and the control is connected to the speaker, display screen and micropump to execute the multidimensional swallowing rehabilitation training method according to any one of claims 1 to 9.