Mistaken swallowing risk identification and early warning method and device for activity-limited persons

Abstract

The invention discloses a mistaken swallowing risk identification early warning method and device for activity-limited persons, and relates to the technical field of risk management, and the method comprises the steps: obtaining the basic physical condition information of a target person, adaptively configuring a swallowing risk monitoring level, and carrying out the initialization setting of a wearable somatosensory device; initial feeding state information is collected for primary risk judgment, if the initial feeding state information passes the primary risk judgment, a feeding permission instruction is issued, real-time feeding state information is collected, emergency risk identification is preferentially carried out, and if no abnormity exists, conventional feeding risk prediction is carried out, and mistaken swallowing risk early warning is carried out. The technical problems that in the prior art, traditional swallowing risk monitoring depends on manpower, subjectivity is high, real-time performance is poor, and accuracy and timeliness of risk identification are low are solved, the wearable somatosensory device and the intelligent data analysis technology are utilized, real-time monitoring and risk early warning of the feeding process are achieved, and the risk early warning efficiency is improved. And the accuracy and timeliness of mistaken swallowing risk identification are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of risk management, and in particular to a method and device for identifying and warning the risk of accidental swallowing for people with limited mobility. Background Art

[0002] With the aging of the population and the increase in the number of patients with chronic diseases (such as stroke, Parkinson's disease, etc.), more and more elderly people and people with limited mobility are at risk of accidental swallowing during eating. Accidental swallowing, that is, food or liquid enters the airway instead of the esophagus, can lead to serious complications such as suffocation, lung infection, and even life-threatening. Therefore, it is particularly important to conduct effective swallowing risk monitoring and early warning for this group of people.

[0003] However, traditional swallowing risk monitoring mainly relies on the experience and observation of nursing staff. However, this method has the problems of strong subjectivity and poor real-time performance, and cannot meet the comprehensive monitoring needs of high-risk populations. Summary of the invention

[0004] The present application provides a method and device for identifying and warning the risk of accidental swallowing for people with limited mobility, which is used to solve the technical problems in the prior art that traditional swallowing risk monitoring relies on manual labor, is highly subjective, has poor real-time performance, and has low accuracy and timeliness in risk identification.

[0005] In a first aspect of the present application, a method for identifying and warning of the risk of accidental swallowing for persons with limited mobility is provided, the method comprising: obtaining basic physical condition information of the target person; adaptively configuring a swallowing risk monitoring level based on the basic physical condition information, and initializing a wearable somatosensory device according to the swallowing risk monitoring level; using the wearable somatosensory device to collect initial eating state information, including food property information and person state information; performing a primary risk determination based on the initial eating state information, including performing a food state risk assessment and a person posture risk assessment, and obtaining a primary risk determination result; if the primary risk determination result is passed, issuing an instruction to allow eating, and using the wearable somatosensory device to collect real-time eating state information of the target person; performing information separation processing on the real-time eating state information to obtain regular eating index information and special eating index information; preferentially performing emergency risk identification based on the special eating index information, and if there is no abnormality in the emergency risk identification result, performing eating risk prediction based on the regular eating index information, and performing an accidental swallowing risk warning based on the eating risk prediction result.

[0006] According to a second aspect of the present application, there is provided an accidental swallowing risk identification and warning device for persons with limited mobility, the device comprising: a person's physical condition information acquisition module, the person's physical condition information acquisition module being used to acquire basic physical condition information of a target person; a device initialization module, the device initialization module being used to adaptively configure a swallowing risk monitoring level based on the basic physical condition information, and to initialize a wearable somatosensory device according to the swallowing risk monitoring level; an initial eating state information acquisition module, the initial eating state information acquisition module being used to use the wearable somatosensory device to collect initial eating state information, including food property information and person status information; a primary risk determination module, the primary risk determination module being used to perform primary risk determination according to the initial eating state information, including initial risk determination; and A food status risk assessment and a personnel posture risk assessment are performed to obtain a primary risk determination result; a real-time eating status information collection module is used to issue an instruction to allow eating if the primary risk determination result is passed, and use the wearable somatosensory device to collect the real-time eating status information of the target person; an information separation processing module is used to separate the real-time eating status information to obtain regular eating index information and special eating index information; an accidental swallowing risk warning module is used to preferentially perform emergency risk identification based on the special eating index information, and if there is no abnormality in the emergency risk identification result, the eating risk is predicted based on the regular eating index information, and an accidental swallowing risk warning is performed based on the eating risk prediction result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The present application provides a method and device for identifying and warning of accidental swallowing risks for persons with limited mobility, which relates to the field of risk management technology. After the wearable somatosensory device is initialized and set based on the basic physical condition information of the target person, the initial eating status information is collected for primary risk judgment. If it passes, an instruction to allow eating is issued, and real-time eating status information is collected, and emergency risk identification is prioritized. If there is no abnormality, routine eating risk prediction is performed, and an accidental swallowing risk warning is issued. The present application solves the technical problems in the prior art that traditional swallowing risk monitoring relies on manual labor, is highly subjective, has poor real-time performance, and has low accuracy and timeliness in risk identification. The present application solves the technical problems that traditional swallowing risk monitoring in the prior art ... has high subjectivity, poor real-time performance, and has low accuracy and timeliness in risk identification. The present application solves the technical problems that traditional swallowing risk monitoring in the prior art relies on manual labor, has high subjectivity, poor real-time performance, and has low accuracy and timeliness in risk identification. The present application solves the technical problems that traditional swallowing risk monitoring in the prior art relies on manual labor, has high subjectivity, poor real-time performance, and has low accuracy and timeliness in risk identification. The present application solves the technical problems that traditional swallowing risk monitoring in the prior art relies on manual labor, BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A schematic diagram of a flow chart of a method for identifying and warning the risk of accidental swallowing for persons with limited mobility provided in an embodiment of the present application;

[0011] Figure 2 A schematic diagram of a process for emergency risk identification based on the special eating index information in a method for identifying and warning the risk of swallowing errors for persons with limited mobility provided in an embodiment of the present application;

[0012] Figure 3 A schematic diagram of the structure of a device for identifying and warning the risk of accidental swallowing for persons with limited mobility provided in an embodiment of the present application.

[0013] Explanation of the accompanying drawings: personnel physical condition information acquisition module 11, equipment initialization module 12, initial eating status information collection module 13, primary risk determination module 14, real-time eating status information collection module 15, information separation and processing module 16, accidental swallowing risk warning module 17. DETAILED DESCRIPTION

[0014] The present application provides a method and device for identifying and warning the risk of accidental swallowing for people with limited mobility, which is used to solve the technical problems in the prior art that traditional swallowing risk monitoring relies on manual labor, is highly subjective, has poor real-time performance, and has low accuracy and timeliness in risk identification.

[0015] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0016] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0017] Embodiment 1, as Figure 1 As shown, the present application provides a method for identifying and warning the risk of accidental swallowing for people with limited mobility, the method comprising:

[0018] P10: Obtain basic physical condition information of the target person.

[0019] Specifically, through the electronic health record system (EHR) or a special medical database, key information such as the target person's medical history, diagnosis report, treatment record, etc. is obtained, and combined with real-time physiological indicator monitoring equipment, the basic physical condition information of the target person is obtained. This information is the basis of the entire risk assessment process and can provide key data support for subsequent adaptive configuration and risk prediction. The basic physical condition information includes but is not limited to the target person's age, medical history, weight and body shape, neurological function status, past accidental swallowing records, etc.

[0020] Among them, age is one of the important factors in measuring swallowing ability. With age, the strength and coordination of throat muscles may weaken, increasing the risk of accidental swallowing. The medical history information of the target person is one of the core elements of the system's assessment of the risk of accidental swallowing. Especially for patients with stroke or Parkinson's disease, these diseases may cause swallowing disorders, making the target person more likely to swallow accidentally when eating. The record of medical history information includes but is not limited to past medical history, current disease status, treatment process, etc. Weight and body shape also affect the safety of swallowing when eating. Being overweight or too thin may change the way food passes through the esophagus, thereby affecting the smoothness of the swallowing process. The system helps assess the stability of the target person's swallowing function by obtaining the target person's body mass index (BMI) and body shape data. The state of neurological function is directly related to the target person's swallowing reflex ability. Diseases or injuries of the nervous system (such as neurodegenerative diseases, brain injuries, etc.) may weaken the swallowing reflex and increase the risk of accidental swallowing. The target person's swallowing reflex can be judged whether it is normal through the evaluation report of the nervous system, and it can be used as an important parameter for risk identification. Furthermore, if the target person has a history of accidental swallowing, the risk monitoring level can be adjusted based on the historical records.

[0021] In addition to the above information, other health data that may affect swallowing safety, such as the health status of the respiratory system and cardiovascular system, can also be collected and analyzed. This information helps to comprehensively assess the health status of the target person and provide support for subsequent risk assessment.

[0022] P20: Based on the basic physical condition information, adaptively configure the swallowing risk monitoring level, and initialize the wearable somatosensory device according to the swallowing risk monitoring level.

[0023] Furthermore, step P20 of the embodiment of the present application further includes:

[0024] P21: According to the user's disease type, multiple key monitoring indicators are configured, and the multiple key monitoring indicators are set with corresponding weight coefficients; P22: Based on the multiple key monitoring indicators, multiple user body indicator values ​​are extracted from the basic physical condition information; P23: The multiple user body indicator values ​​are weighted with reference to the weight coefficients to obtain a comprehensive body indicator value; P24: According to the comprehensive body indicator value, the monitoring level threshold is compared to obtain the swallowing risk monitoring level, and the corresponding somatosensory device parameters are matched according to the swallowing risk monitoring level, and the wearable somatosensory device is initialized and set according to the somatosensory device parameters.

[0025] It should be understood that the swallowing risk monitoring level is adaptively configured based on the user's basic physical condition information, such as age, medical history, current health status, etc., and the monitoring intensity is tailored for the target user to ensure that potential risks are captured in a timely manner while avoiding excessive monitoring. First, multiple key monitoring indicators are configured according to the type of disease of the target person (such as stroke, Parkinson's disease, etc.). These indicators include but are not limited to the number of chewing times, swallowing frequency, breathing rate, head posture stability, etc., and each key monitoring indicator is set with a weight coefficient to reflect its importance in the comprehensive risk assessment. For example, for patients with Parkinson's disease, swallowing frequency and head posture stability may be given higher weight coefficients because these indicators are particularly important in preventing accidental swallowing.

[0026] Furthermore, based on the configured key monitoring indicators, the user's physical index values ​​related to these indicators are extracted from the target person's basic physical condition information. For example, the index value of respiratory rate can be obtained through the respiratory monitoring data in the wearable device. Then, the weighted summation method is used to weight the extracted multiple user physical index values ​​according to their corresponding weight coefficients, and finally a comprehensive physical index value is calculated. This value is a comprehensive quantitative assessment of the user's current swallowing function status, which provides a scientific basis for the subsequent risk level classification.

[0027] Furthermore, the comprehensive physical index value is compared and analyzed with the preset monitoring level threshold. The monitoring level threshold is derived based on a large amount of clinical data and statistical analysis, and is intended to divide users into different risk levels, such as low risk, medium risk, high risk, etc. Based on the comparison results, the user's swallowing risk monitoring level is automatically determined, and the corresponding somatosensory device parameters are matched accordingly. These parameters may include monitoring frequency, sensitivity, alarm threshold, etc., aiming to provide users with a personalized monitoring experience.

[0028] Finally, the wearable somatosensory device is initialized and set according to the matched somatosensory device parameters to ensure that the device can accurately monitor according to the user's specific needs and risk levels, and timely detect and warn of potential swallowing risks.

[0029] P30: Use the wearable somatosensory device to collect initial eating status information, including food property information and personnel status information.

[0030] Optionally, the initial eating status information of the target person is collected through the initialized wearable somatosensory device. This information includes food properties information and personnel status information, which is important basic data for subsequent risk assessment and early warning.

[0031] The food property information refers to the physical and chemical properties of the food that the target person is about to eat, including the texture, size, temperature, humidity, etc. These properties directly affect the difficulty of swallowing by the target person, and thus affect the risk of accidental swallowing.

[0032] Among them, food texture detection includes real-time detection of food hardness and viscosity through pressure sensors installed on dining utensils (such as smart forks and spoons). For patients with dysphagia, foods with higher hardness may increase the difficulty of swallowing, so the system needs to pay special attention to the texture of the food. Size measurement includes photographing the food using a visual sensor with a built-in camera and calculating the volume and shape of the food through an image processing algorithm. For patients who require fine swallowing control, the size of the food chunks is particularly critical, and foods that are too large or irregularly shaped may increase the risk of accidental swallowing. Temperature and humidity detection includes detecting the surface temperature and water content of food through infrared temperature sensors and humidity sensors. Excessively high temperatures may cause burns to the mouth or esophagus, while foods with too low humidity may be difficult to swallow.

[0033] The personnel status information includes the physical and behavioral status of the target person before eating, such as posture, throat muscle activity, breathing rate, etc. This information can help the system determine whether the target person is in a state suitable for eating.

[0034] Among them, posture monitoring includes real-time monitoring of the target person's body posture through accelerometers and gyroscope sensors worn on the head, neck and upper body. The system focuses on the position of the head and neck to ensure that they are maintained at an angle suitable for swallowing when eating. For patients with a higher risk of swallowing, it is crucial to maintain a correct eating posture. Throat electromyography detection includes monitoring the target person's throat muscle activity using an electromyography sensor (EMG) attached to the throat. By analyzing these signals, the system can determine whether the throat muscles are in a suitable state for swallowing, such as whether there is tension or spasm. Respiratory rate monitoring includes recording the target person's breathing rate and depth through a respiratory sensor worn on the chest. By monitoring the breathing pattern, the system can detect abnormal conditions such as shortness of breath and wheezing, which may indicate that the target person is in an unsuitable state before eating.

[0035] Through the above process, the system can comprehensively and accurately obtain the target person's initial eating status information, provide a solid data foundation for subsequent accidental swallowing risk assessment, ensure that the target person is in the best condition before eating, and minimize the risk of accidental swallowing.

[0036] P40: Perform primary risk assessment based on the initial eating status information, including food status risk assessment and personnel posture risk assessment, to obtain a primary risk assessment result.

[0037] Furthermore, step P40 of the embodiment of the present application also includes:

[0038] P41: Based on the initial eating state information, food property information is extracted to perform food property risk assessment, including texture assessment, size assessment, and temperature and humidity assessment, to obtain food property risk assessment results; P42: Based on the initial eating state information, personnel status information is extracted to perform personnel posture risk assessment, including head posture assessment, neck and body posture assessment, and throat electromyography assessment, to obtain personnel posture risk assessment results; P43: Based on the food property risk assessment results and the personnel posture risk assessment results, a comprehensive risk assessment before eating is performed to obtain the primary risk determination result.

[0039] It should be understood that, based on the collected initial eating state information, a primary risk determination is performed to ensure that the target person is in a safe state before eating. This step includes food state risk assessment and person posture risk assessment, and ultimately generates a primary risk determination result.

[0040] First, food property information is extracted from the initial eating state information, and a multi-dimensional risk assessment of the food is performed. This includes texture assessment, size assessment, and temperature and humidity assessment to obtain food property risk assessment results. For example, the hardness and viscosity of the food are detected by the pressure sensor in the smart tableware. Foods that are too hard or too viscous may increase the difficulty of swallowing, especially for target persons with weak swallowing function. These foods may cause food to remain in the mouth or throat, increasing the risk of accidental swallowing. The system marks the food texture as "low risk" or "high risk" based on the detected hardness and viscosity values.

[0041] The size and shape of food are analyzed using visual sensors. The volume and shape of the food block are calculated through image processing algorithms to determine whether it is suitable for the target person to swallow. Large or irregularly shaped food blocks are prone to getting stuck in the throat, increasing the risk of accidental swallowing. The evaluation results are classified according to the suitability of the size, such as "suitable" or "needs adjustment".

[0042] The temperature and humidity of the food surface are detected by the temperature sensor and humidity sensor. If the food temperature is too high, it may cause burns in the mouth or esophagus, and if it is too low, it may cause discomfort during chewing and swallowing. Food with too low humidity may cause dryness during swallowing, increasing the risk. Based on the test results, it is evaluated whether the temperature and humidity of the food are within the safe range.

[0043] After completing the food property assessment, the system will focus on the target person's posture information to ensure that his or her physical condition is suitable for safe eating. This includes head posture assessment, neck and body posture assessment, and throat electromyography assessment to obtain the person's posture risk assessment results. Exemplarily, the system monitors the angle and position of the head in real time through an accelerometer and gyroscope installed on the head. The correct eating posture usually requires the head to be tilted slightly forward to help food pass smoothly through the throat. If the head posture is incorrect (such as leaning back or tilting too much), the system will mark it as "high risk" and prompt the need to adjust the posture.

[0044] The overall posture of the neck and body is monitored through sensors worn on the neck and upper body. Evaluate whether the neck is at the correct angle and whether the body is stable to prevent food from entering the airway due to unstable posture. The posture assessment results will be fed back as "appropriate posture" or "posture needs to be adjusted". Use electromyographic sensors (EMG) in the throat to monitor the activity of throat muscles. Analyze the electromyographic signals to determine whether the target person's throat muscles are ready for swallowing. If abnormal muscle tension or relaxation is detected, the system will mark it as "need to be alert" and prompt the user to make adjustments or further checks.

[0045] Finally, the food properties risk assessment results are integrated with the personnel posture risk assessment results to conduct a comprehensive risk assessment before eating. During this assessment process, the system will combine the two aspects of assessment data to determine whether there is a significant risk of eating under the current conditions.

[0046] If the food properties and posture of the person are within the safe range, the system will generate a "low risk" judgment and issue an instruction to allow eating. If there are some minor risk factors in the assessment (such as slightly high food temperature or slightly deviated posture), the system will generate a "medium risk" judgment and recommend that the user or caregiver make minor adjustments before eating. If there are higher risk factors (such as unpleasant food texture or seriously inappropriate posture), the system will generate a "high risk" judgment, indicating that the eating conditions need to be adjusted or reconfigured immediately.

[0047] P50: If the result of the primary risk assessment is passed, an instruction to allow eating is issued, and the wearable body sensing device is used to collect the target person's real-time eating status information.

[0048] Specifically, if the system draws a passing conclusion based on the results of the primary risk assessment, it means that eating is safe under the current conditions. At this point, the system will issue an instruction to allow eating, and at the same time start the real-time monitoring function of the wearable somatosensory device to continuously collect the target person's real-time eating status information to further ensure the safety of the eating process. The real-time eating status information includes various physiological and behavioral data related to swallowing safety of the target person during the eating process, such as chewing frequency, swallowing movements, breathing patterns, and head posture. For example, the chewing frequency of the target person can be monitored in real time through an electromyographic sensor installed near the chewing muscles. The time and frequency of each swallowing action can be monitored through the throat electromyographic sensor and motion capture equipment.

[0049] P60: Separating the real-time eating status information to obtain regular eating index information and special eating index information.

[0050] In a possible embodiment of the present application, the collected real-time eating status information is processed for information separation to distinguish between regular eating index information and special eating index information, and accurately identify the target person's normal eating behavior and potential risk behavior, thereby providing reliable data support for subsequent risk assessment and early warning.

[0051] First, the real-time eating status information is preprocessed, including denoising, filtering, data cleaning, etc., to improve data quality and reduce interference factors. Subsequently, feature extraction algorithms, such as fast Fourier transform (FFT) and wavelet transform, are used to perform frequency domain analysis and time domain analysis of the data to accurately extract characteristic signals of behaviors such as chewing, swallowing, and breathing, and identify and separate regular eating indicator information from the preprocessed data. This type of information usually reflects the physiological characteristics and behavioral patterns that are common in most people during normal eating, such as chewing frequency, swallowing frequency, breathing frequency, head posture, etc., which can be used to evaluate the overall state and efficiency of the eating process.

[0052] At the same time, the system also pays close attention to eating indicators that may be beyond the normal range or have special significance. These special eating indicators are often associated with specific health problems or risk conditions, such as abnormal swallowing movements, sudden choking reactions, and abnormal changes in throat myoelectric activity. For example, sudden muscle activity or changes in breathing patterns detected by throat myoelectric sensors and respiratory sensors, and abnormal breathing patterns detected by respiratory sensors, such as rapid breathing and short pauses, can help the system identify potential risks of accidental swallowing in a timely manner.

[0053] P70: Emergency risk identification is performed preferentially based on the special eating indicator information. If there is no abnormality in the emergency risk identification result, eating risk prediction is performed based on the conventional eating indicator information, and an accidental swallowing risk warning is issued based on the eating risk prediction result.

[0054] Specifically, during real-time monitoring, priority is given to marked special eating indicator information. This information includes behaviors or physiological reactions that may indicate high risks, such as coughing, shortness of breath, and uncoordinated swallowing. Since these special indicators are often directly related to emergency situations, the system regards them as core data for emergency risk identification. If the emergency risk identification results show an abnormality, the system will immediately issue an early warning signal. Early warnings can be conveyed in a variety of ways, such as sound alarms, equipment vibrations, or visual prompts, to remind the target person or caregiver to take immediate intervention measures, such as pausing eating, adjusting posture, or seeking medical help.

[0055] If the emergency risk identification results show that there are no abnormalities, the system will continue to conduct in-depth analysis of the target person's regular eating indicator information. At this point, the system will no longer focus on emergency situations, but will predict potential eating risks through more extensive data analysis. Perform a multi-dimensional comprehensive analysis of the extracted regular eating indicator information. Including the integration and evaluation of various data such as chewing frequency, swallowing frequency, breathing pattern and head posture. By comparing and analyzing the historical data and real-time data of these indicators, we can identify deviations and abnormalities that may occur during the eating process, obtain eating risk prediction results, and decide whether to trigger an accidental swallowing risk warning based on the results of the eating risk prediction.

[0056] Further, such as Figure 2 As shown, emergency risk identification is performed based on the special eating indicator information, and step P70 of the embodiment of the present application also includes:

[0057] P71: Based on the accidental swallowing experience data, multiple special eating indicators are extracted, and corresponding emergency risk identification channels are configured respectively; P72: Based on the special eating indicator information, the target emergency risk identification channel is matched, and emergency risk identification is performed according to the target emergency risk identification channel to obtain the emergency risk identification result.

[0058] Optionally, a number of key special eating indicators are extracted based on a large amount of experience data on accidental swallowing. These indicators include, but are not limited to, food texture (such as liquid, semi-liquid or solid), eating speed, coordination of swallowing movements, physiological reactions during eating (such as coughing, changes in complexion), and patients' specific health conditions (such as cognitive impairment, history of dysphagia). Each special eating indicator is considered an important factor in assessing potential risks during the eating process, and is therefore configured with an independent emergency risk identification channel. These channels are like preset warning lines, each of which accurately monitors and evaluates different types of risk characteristics. For example, the cough recognition channel specifically monitors and analyzes changes in cough frequency and intensity. Once an abnormal cough is detected, the channel will initiate an emergency response. The breathing recognition channel monitors changes in breathing patterns in real time. When rapid or irregular breathing is detected, the channel will trigger an alarm.

[0059] During real-time monitoring, when the system receives special eating indicator information of the target person, it matches the most relevant target emergency risk identification channel based on this information. By comparing the real-time collected data with the preset indicator characteristics, the appropriate identification channel is selected. For example, if frequent coughing signals are detected, the cough identification channel is automatically activated for further analysis.

[0060] Once the target emergency risk identification channel is activated, the system immediately performs emergency risk identification. The identification process includes an in-depth analysis of the current special eating indicators to determine whether there is significant swallowing or other emergency situations. For example: in the cough identification channel, the frequency, intensity and duration of the cough are analyzed to determine whether it is a normal cough or a dangerous cough caused by swallowing. In the breathing identification channel, changes in breathing patterns are evaluated to determine whether shortness of breath is accompanied by other dangerous signals (such as simultaneous posture changes).

[0061] The recognition results of each channel are integrated to generate a comprehensive emergency risk recognition result. If the recognition result of any channel shows a high risk, the system will immediately trigger the corresponding early warning mechanism to notify the target person or nursing staff to take emergency measures.

[0062] Furthermore, if there is no abnormality in the emergency risk identification result, step P70 of the embodiment of the present application further includes:

[0063] P73: The conventional eating index information includes but is not limited to the number of chewing times, swallowing frequency, breathing frequency and head posture stability; P74: Based on the conventional eating index information, a conventional index matrix is ​​constructed: Among them, X is the conventional indicator matrix, n is the number of samples, m is the number of indicators, and x ij Represents the value of the i-th sample on the j-th indicator; P75: Normalize the conventional indicator matrix to obtain the standard indicator matrix: Among them, R is the standard indicator matrix, r ij is the standardized value; P76: Calculate and obtain the ideal solution and negative ideal solution of each regular eating index, wherein the ideal solution is: A + ={r1 + ,r2 + ,…,r m +}; The negative ideal solution is: A - ={r1 - ,r2 - ,…,r m -}; P77: Based on the ideal solution and the negative ideal solution, combined with the standard indicator matrix, eating risk prediction is performed to obtain an eating risk prediction result.

[0064] It should be understood that if there is no abnormality in the emergency risk identification result, the system will continue to analyze the regular eating index information to predict the potential risks of the target person during eating. The regular eating index information includes but is not limited to the number of chewing times, swallowing frequency, breathing frequency and head posture stability. Based on the collected regular eating index information, an n×m regular index matrix is ​​constructed: Among them, X is the conventional indicator matrix, n is the number of samples, m is the number of indicators, and x ij Indicates the value of the i-th sample on the j-th index.

[0065] In order to eliminate the dimensional differences between different indicators, the system normalizes the conventional indicator matrix and standardizes the value of each indicator into a dimensionless form. The standardized matrix is ​​called the standard indicator matrix, and its form is: Among them, R is the standard indicator matrix, and the formula for standardization is: Among them, r ij is the standardized value.

[0066] Furthermore, the system calculates the ideal solution (optimal solution) and negative ideal solution (worst solution) of each regular eating indicator based on the standardized indicator value. The calculation of the ideal solution and negative ideal solution is as follows: The ideal solution is: A + ={r1 + ,r2 + ,…,r m +}; Ideal solution A + It represents the value of each indicator in the best state, corresponding to the maximum or minimum value, depending on the nature of the indicator. The negative ideal solution is: A - ={r1 - ,r2 - ,…,r m -}; A- Indicates the value of each indicator in the worst state. The calculation basis of the ideal solution and the negative ideal solution is the extreme case of the standardized value of each indicator, which can be used to measure the distance between the current eating state and the ideal state or the worst state.

[0067] Finally, the standardized values ​​in the standard indicator matrix are used to combine the ideal solution and the negative ideal solution to predict the risk of eating. The core of risk prediction is to calculate the Euclidean distance of each sample to the ideal solution and the negative ideal solution, and evaluate the safety of the eating state based on this to obtain the eating risk prediction result.

[0068] Furthermore, step P77 of the embodiment of the present application also includes:

[0069] P77-1: Based on the standard indicator matrix, calculate the distance between each sample and the ideal solution and the negative ideal solution to obtain the sample relative distance. The formula is:

[0070] in, is the relative distance of the positive samples, is the negative sample relative distance; P77-2: Based on the sample relative distance, calculate the sample relative proximity: Among them, C i is the relative proximity of the samples, which is between 0 and 1; P77-3: performing eating risk prediction based on the relative proximity of the samples, and performing accidental swallowing risk warning based on the eating risk prediction results.

[0071] Optionally, first, the system calculates the distance between each sample and the ideal solution and the negative ideal solution based on the previously constructed standard indicator matrix. For each sample, its distance from the ideal solution is calculated, which is called the positive sample relative distance, and the formula is as follows: in, is the relative distance of the positive sample, indicating the distance from the i-th sample to the ideal solution A + The Euclidean distance of . Ideal solution A + represents the value when all indicators are in the best state. At the same time, the distance between each sample and the negative ideal solution is calculated, which is called the negative sample relative distance: A - ={r1 - ,r2 - ,…,r m -}; A - Indicates the value of each indicator in the worst state.

[0072] After calculating the relative distance of the positive sample Relative distance to negative samples Finally, the system evaluates the risk of each sample by calculating the relative proximity of the samples. The calculation formula for the relative proximity of the samples is as follows: Among them, C i is the relative proximity of the sample, which is between 0 and 1 and is used to measure the degree of proximity of the sample to the ideal solution. The closer the value is to 1, the closer the sample is to the ideal state and the lower the eating risk; the closer the value is to 0, the closer the sample is to the negative ideal state and the higher the eating risk. The eating risk is predicted based on the relative proximity of the sample, and the risk of swallowing error is warned based on the eating risk prediction result.

[0073] Furthermore, step P77-3 of the embodiment of the present application also includes:

[0074] P77-31: Preset the risk level division threshold; P77-32: According to the relative proximity of the samples, compare the risk level division threshold, perform risk level division, and obtain the eating risk prediction result; P77-33: According to the eating risk prediction result, match the target warning mechanism to issue an accidental swallowing risk warning.

[0075] Specifically, firstly, based on historical data and actual eating monitoring experience, the risk level classification thresholds are preset. These thresholds are used to classify the relative proximity of samples into different risk levels. For example, the low risk threshold is usually set at the proximity C i ≥0.8, indicating that eating behavior is in an ideal state and the risk of accidental swallowing is extremely low. The medium risk threshold is set at 0.5≤C i <0.8, indicating that there is a certain risk of eating, but the risk is controllable. The high risk threshold is set at C i <0.5 indicates that there is a significant risk of accidental swallowing during eating, and immediate measures need to be taken. These risk thresholds can be continuously optimized according to actual conditions to improve the accuracy of risk prediction.

[0076] The relative proximity of each sample is compared with the preset risk level threshold to determine the corresponding risk level. i ≥0.8, the sample is classified as low risk, indicating that the target person's eating behavior is very safe and no special intervention is required. i <0.8, the sample is classified as medium risk, indicating that there may be some factors that need attention during eating, such as unstable posture or insufficient chewing. i <0.5, the sample is classified as high risk, and the system identifies that the target person may be at a high risk of accidental swallowing and requires immediate intervention.

[0077] According to the risk level classification, the eating risk prediction results are generated, and according to the eating risk prediction results, the appropriate target early warning mechanism is matched to issue accidental swallowing risk warning. Different risk levels correspond to different warning intensities and methods. For example, when the risk is low, the system may only give slight prompts, such as visual or audio reminders, to inform the target person that the current eating state is good and to continue. When the risk is medium, the system will issue a clear warning signal, suggesting that the target person or caregiver pay attention to some potential problems during the eating process and adjust the eating posture or rhythm if necessary. When the risk is high, the system will trigger a strong early warning mechanism, such as equipment vibration, sound alarm, or immediately notify the caregiver to stop eating, check the status of the food or the target person, and prevent accidental swallowing accidents.

[0078] In summary, the embodiments of the present application have at least the following technical effects:

[0079] This application uses the basic physical condition information of the target person to adaptively configure the swallowing risk monitoring level. After initializing the wearable somatosensory device, it collects the initial eating status information for primary risk judgment. If passed, it issues an instruction to allow eating, and collects real-time eating status information, giving priority to emergency risk identification. If there is no abnormality, it conducts routine eating risk prediction and conducts accidental swallowing risk warning.

[0080] The technical effect of utilizing wearable somatosensory devices and intelligent data analysis technology to achieve real-time monitoring and risk warning of the eating process, and to improve the accuracy and timeliness of identifying the risk of accidental swallowing, has been achieved.

[0081] Embodiment 2, based on the same inventive concept as the method for identifying and warning the risk of accidental swallowing for persons with limited mobility in the aforementioned embodiment, Figure 3 As shown, the present application provides a device for identifying and warning the risk of accidental swallowing for people with limited mobility. The device and method embodiments in the present application are based on the same inventive concept. The device includes:

[0082] The personnel physical condition information acquisition module 11 is used to acquire basic physical condition information of a target person.

[0083] The device initialization module 12 is used to adaptively configure the swallowing risk monitoring level based on the basic physical condition information, and initialize the wearable somatosensory device according to the swallowing risk monitoring level.

[0084] The initial eating state information collection module 13 is used to use the wearable somatosensory device to collect initial eating state information, including food property information and personnel status information.

[0085] The primary risk determination module 14 is used to perform primary risk determination according to the initial eating state information, including food state risk assessment and personnel posture risk assessment, to obtain a primary risk determination result.

[0086] The real-time eating status information collection module 15 is used to issue an instruction to allow eating if the primary risk determination result is passed, and use the wearable somatosensory device to collect the real-time eating status information of the target person.

[0087] The information separation processing module 16 is used to perform information separation processing on the real-time eating status information to obtain regular eating index information and special eating index information.

[0088] The accidental swallowing risk warning module 17 is used to preferentially identify emergency risks based on the special eating index information. If there is no abnormality in the emergency risk identification result, eating risk prediction is performed based on the conventional eating index information, and an accidental swallowing risk warning is performed based on the eating risk prediction result.

[0089] Furthermore, the device initialization module 12 is further configured to perform the following steps:

[0090] According to the user's disease type, multiple key monitoring indicators are configured, and the multiple key monitoring indicators are set with corresponding weight coefficients; based on the multiple key monitoring indicators, multiple user body indicator values ​​are extracted from the basic physical condition information; the multiple user body indicator values ​​are weighted with reference to the weight coefficients to obtain a comprehensive body indicator value; according to the comprehensive body indicator value, the monitoring level threshold is compared to obtain the swallowing risk monitoring level, and the corresponding somatosensory device parameters are matched according to the swallowing risk monitoring level, and the wearable somatosensory device is initialized and set according to the somatosensory device parameters.

[0091] Furthermore, the primary risk determination module 14 is further configured to perform the following steps:

[0092] Based on the initial eating state information, food property information is extracted to perform food property risk assessment, including texture assessment, size assessment, and temperature and humidity assessment, to obtain food property risk assessment results; based on the initial eating state information, personnel status information is extracted to perform personnel posture risk assessment, including head posture assessment, neck and body posture assessment, and throat electromyography assessment, to obtain personnel posture risk assessment results; based on the food property risk assessment results and the personnel posture risk assessment results, a comprehensive risk assessment before eating is performed to obtain the primary risk determination result.

[0093] Furthermore, the accidental swallowing risk warning module 17 is further configured to perform the following steps:

[0094] According to the accidental swallowing experience data, multiple special eating indicators are extracted, and corresponding emergency risk identification channels are configured respectively; based on the special eating indicator information, the target emergency risk identification channel is matched, and emergency risk identification is performed according to the target emergency risk identification channel to obtain the emergency risk identification result.

[0095] Furthermore, the accidental swallowing risk warning module 17 is further configured to perform the following steps:

[0096] The conventional eating index information includes but is not limited to the number of chewing times, swallowing frequency, breathing frequency and head posture stability; based on the conventional eating index information, a conventional index matrix is ​​constructed: Among them, X is the conventional indicator matrix, n is the number of samples, m is the number of indicators, and x ij represents the value of the i-th sample on the j-th indicator; the conventional indicator matrix is ​​normalized to obtain the standard indicator matrix: Among them, R is the standard indicator matrix, r ij is the standardized value; calculate and obtain the ideal solution and negative ideal solution of each regular eating index, wherein the ideal solution is: A + ={r1 + ,r2 + ,…,r m +}; The negative ideal solution is: A - ={r1 - ,r2 - ,…,r m -}; Based on the ideal solution and the negative ideal solution, combined with the standard indicator matrix, eating risk prediction is performed to obtain an eating risk prediction result.

[0097] Furthermore, the accidental swallowing risk warning module 17 is further configured to perform the following steps:

[0098] Based on the standard indicator matrix, the distance between each sample and the ideal solution and the negative ideal solution is calculated to obtain the sample relative distance, and the formula is: in, is the relative distance of the positive samples, is the negative sample relative distance; based on the sample relative distance, the sample relative proximity is calculated: Among them, C i is the relative proximity of the samples, which is between 0 and 1; an eating risk prediction is performed based on the relative proximity of the samples, and an accidental swallowing risk warning is performed based on the eating risk prediction result.

[0099] Furthermore, the accidental swallowing risk warning module 17 is further configured to perform the following steps:

[0100] Preset a risk level division threshold; perform risk level division based on the relative proximity of the samples and compare the risk level division threshold to obtain the eating risk prediction result; and match the target warning mechanism to perform accidental swallowing risk warning based on the eating risk prediction result.

[0101] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0102] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0103] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for identifying and warning the risk of accidental swallowing for people with limited mobility, characterized in that: The method comprises: Obtain basic physical condition information of the target person; Based on the basic physical condition information, adaptively configure a swallowing risk monitoring level, and initialize and set the wearable somatosensory device according to the swallowing risk monitoring level; Using the wearable somatosensory device, collecting initial eating status information, including food property information and personnel status information; Performing a primary risk assessment based on the initial eating state information, including performing a food state risk assessment and a personnel posture risk assessment, to obtain a primary risk assessment result; If the primary risk determination result is passed, an instruction to allow eating is issued, and the wearable body sensing device is used to collect the target person's real-time eating status information; Separating the real-time eating status information to obtain regular eating index information and special eating index information; Emergency risk identification is preferentially performed based on the special eating indicator information. If there is no abnormality in the emergency risk identification result, eating risk prediction is performed based on the conventional eating indicator information, and an accidental swallowing risk warning is performed based on the eating risk prediction result.

2. The method for identifying and warning the risk of accidental swallowing for people with limited mobility as claimed in claim 1, characterized in that: Based on the basic physical condition information, adaptively configuring the swallowing risk monitoring level, and initializing the wearable somatosensory device according to the swallowing risk monitoring level, including: According to the user's disease type, multiple key monitoring indicators are configured, and the multiple key monitoring indicators are set with corresponding weight coefficients; Based on the multiple key monitoring indicators, extracting multiple user physical indicator values ​​from the basic physical condition information; The multiple user body index values ​​are weighted with reference to the weight coefficient to obtain a comprehensive body index value; According to the comprehensive physical indicator value, the monitoring level threshold is compared to obtain the swallowing risk monitoring level, and the corresponding somatosensory device parameters are matched according to the swallowing risk monitoring level, and the wearable somatosensory device is initialized and set according to the somatosensory device parameters.

3. The method for identifying and warning the risk of accidental swallowing for people with limited mobility as claimed in claim 1, characterized in that: A primary risk determination is performed based on the initial eating state information, including food state risk assessment and personnel posture risk assessment, including: Based on the initial eating state information, extracting food property information to perform food property risk assessment, including texture assessment, size assessment, and temperature and humidity assessment, to obtain a food property risk assessment result; Based on the initial eating state information, extracting personnel state information, performing personnel posture risk assessment, including head posture assessment, neck and body posture assessment, and throat electromyography assessment, to obtain personnel posture risk assessment results; Based on the food property risk assessment results and the personnel posture risk assessment results, a comprehensive risk assessment before eating is performed to obtain the primary risk determination result.

4. The method for identifying and warning the risk of accidental swallowing for people with limited mobility as claimed in claim 1, characterized in that: Emergency risk identification is performed based on the special eating indicator information, including: Based on the experience data of swallowing errors, multiple special eating indicators are extracted and corresponding emergency risk identification channels are configured respectively; Based on the special eating indicator information, a target emergency risk identification channel is matched, and emergency risk identification is performed according to the target emergency risk identification channel to obtain the emergency risk identification result.

5. The method for identifying and warning the risk of accidental swallowing for people with limited mobility as claimed in claim 1, characterized in that: If the emergency risk identification result is normal, the eating risk prediction is performed based on the conventional eating index information, including: The conventional eating index information includes but is not limited to the number of chewing times, swallowing frequency, breathing frequency and head posture stability; Based on the conventional eating index information, a conventional index matrix is ​​constructed: Among them, X is the conventional indicator matrix, n is the number of samples, m is the number of indicators, and x ij Indicates the value of the i-th sample on the j-th index; The conventional indicator matrix is ​​normalized to obtain a standard indicator matrix: Among them, R is the standard indicator matrix, r ij is the standardized value; Calculate and obtain the ideal solution and negative ideal solution of each regular eating index, where the ideal solution is: A + ={r1 + ,r2 + ,…,r m + }; The negative ideal solution is: A - ={r1 - ,r2 - ,…,r m - }; Based on the ideal solution and the negative ideal solution, eating risk prediction is performed in combination with the standard indicator matrix to obtain an eating risk prediction result.

6. The method for identifying and warning the risk of accidental swallowing for people with limited mobility as claimed in claim 5, characterized in that: Based on the ideal solution and the negative ideal solution, combined with the standard indicator matrix, the eating risk prediction is performed to obtain the eating risk prediction result, including: Based on the standard indicator matrix, the distance between each sample and the ideal solution and the negative ideal solution is calculated to obtain the sample relative distance. The formula is: in, is the relative distance of the positive samples, is the relative distance of negative samples; Based on the sample relative distance, calculate the sample relative proximity: Among them, C i is the relative proximity of the samples, between 0 and 1; An eating risk prediction is performed based on the relative proximity of the samples, and an accidental swallowing risk warning is issued based on the eating risk prediction result.

7. The method for identifying and warning the risk of accidental swallowing for persons with limited mobility as claimed in claim 6, characterized in that: The eating risk is predicted based on the relative proximity of the samples, and the risk of accidental swallowing is warned based on the eating risk prediction result, including: Preset risk level classification thresholds; According to the relative proximity of the samples, the risk level classification threshold is compared to perform risk level classification, and the eating risk prediction result is obtained; According to the eating risk prediction result, the target warning mechanism is matched to perform accidental swallowing risk warning.

8. A device for identifying and warning the risk of accidental swallowing for people with limited mobility, characterized in that: The device comprises: A personnel physical condition information acquisition module, which is used to acquire basic physical condition information of a target person; A device initialization module, the device initialization module is used to adaptively configure the swallowing risk monitoring level based on the basic physical condition information, and initialize the wearable somatosensory device according to the swallowing risk monitoring level; An initial eating state information collection module, the initial eating state information collection module is used to use the wearable somatosensory device to collect initial eating state information, including food property information and personnel status information; A primary risk determination module, which is used to perform primary risk determination according to the initial eating state information, including performing food state risk assessment and personnel posture risk assessment, and obtaining a primary risk determination result; A real-time eating status information collection module, wherein the real-time eating status information collection module is used to issue an instruction to allow eating if the primary risk determination result is passed, and use the wearable somatosensory device to collect the real-time eating status information of the target person; An information separation processing module, the information separation processing module is used to perform information separation processing on the real-time eating status information to obtain regular eating index information and special eating index information; The accidental swallowing risk warning module is used to preferentially identify emergency risks based on the special eating index information. If there is no abnormality in the emergency risk identification result, eating risk prediction is performed based on the conventional eating index information, and an accidental swallowing risk warning is performed based on the eating risk prediction result.

Images