Personalized upper airway rehabilitation training method based on osa severity grading

Abstract

The application relates to the field of medical surgical technology, in particular to a personalized upper airway rehabilitation training method based on OSA severity grading, which comprises the following steps: determining a preliminary rehabilitation training scheme according to the OSA severity of a target user; determining optimized physiological indexes and unoptimized physiological indexes; determining a first similar user of the target user; a characteristic vector comprises the unoptimized physiological indexes and a change trend in the preliminary rehabilitation process; determining an updated item of the target user according to the correlation between the physiological indexes of the first similar user when performing a historical rehabilitation training item and the historical rehabilitation training item performed by the first similar user; adding the updated item into the preliminary rehabilitation training scheme of the target user to obtain an updated rehabilitation training scheme; and performing rehabilitation training on the upper airway of the target user by using the updated rehabilitation training scheme. Thus, the rehabilitation training effect and safety of the patient are improved.

Description

Technical Field

[0001] This invention relates to the field of surgical medical technology, specifically to a personalized upper airway rehabilitation training method based on OSA severity grading. Background Technology

[0002] Obstructive sleep apnea-hypopnea syndrome (OSA) is a common sleep-disordered breathing disorder characterized by recurrent upper airway collapse and obstruction during sleep. It can lead to nocturnal hypoxemia, sleep structure disturbances, and other problems, severely impacting patients' sleep quality and overall health. Currently, OSA severity grading is the basis for developing rehabilitation training programs, and rehabilitation interventions are often directly tailored to the patient's OSA grade, matching a general training program to that grade. However, clinical practice has revealed significant individual differences in upper airway anatomy, collapse mechanisms, and pathological characteristics among patients with the same OSA grade. Some patients develop the condition primarily due to upper airway soft tissue relaxation, while others develop it due to weak peripharyngeal muscles or abnormal respiratory regulation. Even with the same OSA grade, the pathogenesis and pathological basis differ.

[0003] In some scenarios, existing rehabilitation training programs only use the severity of OSA as a single criterion, without fully considering the differences among patients in terms of upper airway collapse mechanisms, muscle function status, and respiratory regulation characteristics. This results in insufficient matching between the rehabilitation training program and the patient's actual pathological characteristics, poor targeting and effectiveness of rehabilitation training, and consequently, poor rehabilitation training results and safety for patients. Summary of the Invention

[0004] To address the technical problem of poor rehabilitation training effectiveness and safety for patients, the present invention aims to provide a personalized upper airway rehabilitation training method based on OSA severity grading.

[0005] To solve the above technical problems, the specific technical solution adopted is as follows: In a first aspect, embodiments of the present invention provide a personalized upper airway rehabilitation training method based on OSA severity grading, comprising: determining a preliminary rehabilitation training plan based on the OSA severity of the target user; determining optimized and unoptimized physiological indicators based on various physiological indicators of the target user before and during the preliminary rehabilitation training plan; determining a first similar user of the target user based on the similarity between a first feature vector of the target user and a second feature vector of historical users, wherein the feature vector includes unoptimized physiological indicators and trends of change during the preliminary rehabilitation process; determining updated items for the target user based on the correlation between the physiological indicators of the first similar user when performing historical rehabilitation training items and the historical rehabilitation training items performed by the first similar user; adding the updated items to the preliminary rehabilitation training plan of the target user to obtain an updated rehabilitation training plan, and using the updated rehabilitation training plan to perform upper airway rehabilitation training on the target user.

[0006] Optionally, determining optimized and non-optimized physiological indicators based on the target user's physiological indicators before and during the initial rehabilitation training program includes: normalizing the target user's physiological indicators before and during the initial rehabilitation training program to obtain normalized values; performing linear fitting on the normalized values ​​before and during the initial rehabilitation training program to obtain the first trend of change of each physiological indicator before the initial rehabilitation training program and the second trend of change of each physiological indicator during the initial rehabilitation program; determining improvement indicators from the physiological indicators based on the difference between the second and first trends of change of the physiological indicators; determining the improvement enhancement state of the improvement indicators during the initial rehabilitation program using the difference between the second and first trends of change of the improvement indicators and the first trend of change; and determining optimized physiological indicators that meet the physiological state optimization conditions and non-optimized physiological indicators that do not meet the physiological state optimization conditions based on the difference between the second and first trends of change of the improvement indicators, the improvement enhancement state, and the normalized values ​​of the improvement indicators during the initial rehabilitation program.

[0007] Optionally, the physiological state optimization conditions include: the normalized value of the product of the difference between the second and first trends of the improvement indicators and the improvement enhancement state is greater than a preset value; or the average value of the normalized values ​​of the improvement indicators during the initial recovery process is within the normal range of physiological state values.

[0008] Optionally, determining the first similar user of the target user based on the similarity between the first feature vector of the target user and the second feature vector of historical users includes: determining the second similar user of the target user based on the cosine similarity between the first feature vector of the target user and the second feature vector of historical users; and selecting the first similar user from the second similar users based on the difference between the second and first trends of the optimized physiological indicators of the first similar user and the corresponding improvement and enhancement state, and the difference between the second and first trends of the optimized physiological indicators of the target user and the corresponding improvement and enhancement state.

[0009] Optionally, selecting first similar users from second similar users based on the difference between the second and first trends of the optimized physiological indicators of the first similar users and the corresponding improvement and enhancement states, and the difference between the second and first trends of the optimized physiological indicators of the target users and the corresponding improvement and enhancement states, includes selecting first similar users from second similar users whose differences between the second and first trends of the optimized physiological indicators and the corresponding improvement and enhancement states are both greater than those of the target users.

[0010] Optionally, determining the updated items for the target user based on the correlation between the physiological indicators of the first similar user during historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user includes: obtaining the rehabilitation training programs of the first similar user; obtaining the correlation matrix between the normalized values ​​of the physiological indicators of the first similar user during rehabilitation training programs and the rehabilitation training programs performed by the first similar user; selecting target rehabilitation training programs from the correlation matrix where the correlation between the normalized values ​​of the unoptimized physiological indicators of the target user and the rehabilitation training programs performed by the target user is greater than a preset threshold; determining the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators based on the correlation between the target user's current target rehabilitation training program and the unoptimized and optimized physiological indicators and the program's experience value; and determining the updated items based on the effectiveness score.

[0011] Optionally, obtaining the correlation matrix between the normalized values ​​of the physiological indicators of the first similar users when they perform rehabilitation training programs and the rehabilitation training programs performed by the first similar users includes: determining a first average value of the training duration of each first similar user's rehabilitation training program and a second average value of the normalized values ​​of the physiological indicators of the first similar users when they perform rehabilitation training programs; calculating a first difference between the training duration of each first similar user's rehabilitation training program and the first average value, and a second difference between the normalized values ​​of the physiological indicators of the first similar users' rehabilitation training programs and the second average value; and determining the correlation matrix between each rehabilitation training program and the physiological indicators based on the first difference and the second difference.

[0012] Optionally, the method for determining the project experience value includes: the ratio between the first number of first similar users corresponding to the target rehabilitation training project currently being performed by the target user and the second number of all first similar users.

[0013] Optionally, determining the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators based on the correlation between the target rehabilitation training program currently being performed by the target user and the unoptimized and optimized physiological indicators and the program experience value includes: determining the product of the correlation between the target rehabilitation training program currently being performed by the target user and the unoptimized physiological indicators and the program experience value as the first effective score; The second effective score is determined by multiplying the correlation between the target rehabilitation training program currently being performed by the target user and the optimized physiological indicators with the program's experience value. The effectiveness score includes the first effective score and the second effective score.

[0014] Optionally, determining the update items based on the validity score includes: sorting the first valid scores in descending order, and selecting the top N items with the first valid scores greater than the validity threshold and the highest ranking as the first update items; sorting the second valid scores in descending order, and selecting the top M items with the highest ranking as the second update items, where M is greater than N, and the update items include the first update items and the second update items.

[0015] This invention offers the following advantages: First, it determines an initial rehabilitation plan based on the severity of OSA. Then, relying on the optimization of the patient's individual physiological indicators and the trends in indicator changes during rehabilitation, it identifies unimproved physiological indicators and weaknesses in rehabilitation. This allows for a training plan more closely aligned with the patient's unique pathological characteristics and physiological state, ensuring the rehabilitation plan is tailored to the patient's actual condition and improving its suitability. Furthermore, by comparing various physiological indicators before and after initial rehabilitation training, and during the rehabilitation process, it distinguishes between optimized and unoptimized physiological indicators, identifying weaknesses in the patient's rehabilitation training and avoiding ineffective or redundant training programs. Simultaneously, by combining the similarity between the patient's feature vector and historical user feature vectors, it filters similar users with high matching degrees. Based on rehabilitation data from similar historical cases, it mines the correlation between physiological indicators and training programs, selecting suitable and efficient training programs for target users, taking into account individual differences and clinical patterns, making the development of rehabilitation plans based on evidence. Targeted supplementation with updated training programs effectively compensates for the shortcomings of the initial training plan, making rehabilitation training more targeted and significantly improving training effectiveness. Therefore, the embodiments of the present invention, through personalized optimization and updating of training programs, conduct training in accordance with the patient's actual physiological tolerance and pathological needs. This can not only accurately improve problems such as upper airway collapse, abnormal muscle function, and respiratory regulation disorder, and steadily improve the rehabilitation effect, but also avoid unsuitable training programs that would increase the patient's physical burden, thereby improving rehabilitation efficiency and training safety, and comprehensively improving the overall quality of rehabilitation training. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart illustrating a personalized upper airway rehabilitation training method based on OSA severity grading, provided as an embodiment of the present invention; Figure 2 This is a schematic diagram of a preliminary rehabilitation training program provided in one embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a personalized upper airway rehabilitation training method based on OSA severity grading proposed according to the present invention.

[0019] The specific scenario targeted by this invention is: when patients undergo sleep apnea monitoring and OSA rehabilitation training in a home or medical institution environment, especially after the monitoring results have made a preliminary classification of the severity of OSA, it is necessary to dynamically adjust the rehabilitation training program for upper airway muscle training for patients with different upper airway collapse mechanisms and respiratory function characteristics, so as to evaluate and optimize the suitability of rehabilitation training.

[0020] The following description, in conjunction with the accompanying drawings, details a specific scheme for a personalized upper airway rehabilitation training method based on OSA severity grading provided by this invention.

[0021] Example 1: Please see Figure 1 The diagram illustrates a flowchart of a personalized upper airway rehabilitation training method based on OSA severity grading provided in an embodiment of the present invention, including: Step S101: Determine a preliminary rehabilitation training plan based on the severity of OSA in the target user.

[0022] Specifically, in this embodiment of the invention, when obtaining the severity of a patient's OSA, it is generally calculated using the Obstructive Sleep Apnea Index (AHI). The AHI represents the total number of apnea and hypopnea events occurring within a unit of sleep time, reflecting the frequency of respiratory abnormalities during sleep. Based on the AHI value range, the severity of the patient's OSA is divided into different levels, including: Normal state: times / hour, mild OSA: Times / hour, moderate OSA: Times / hour, severe OSA: times / hour.

[0023] Furthermore, exemplaryly, embodiments of the present invention configure the following based on the severity of the patient's OSA: Figure 2 The corresponding preliminary rehabilitation training program is shown. Figure 2 This is a schematic diagram of a preliminary rehabilitation training program provided in an embodiment of the present invention. It is worth noting that, depending on the actual situation, the preliminary rehabilitation training program can be of other types, and this embodiment of the present invention does not limit it.

[0024] Furthermore, to develop personalized rehabilitation training programs for different patients, it is necessary to comprehensively analyze the patient's sleep breathing and blood oxygenation status. For example, this invention selects four physiological indicators—AHI, percentage of hypoxic time, lowest blood oxygen saturation, and nocturnal respiratory instability index—as indicators to analyze the suitability of the rehabilitation training program for the patient. The specific process is as follows: Determine the current patient's AHI and assess OSA severity: Obtain the patient's nighttime sleep data through a wearable device and calculate the AHI. Based on the AHI value range, classify the patient's OSA severity into different levels. Record the current patient's physical indicators and determine the initial rehabilitation training program: Simultaneously collect the patient's percentage of hypoxic time, lowest blood oxygen saturation, and nocturnal respiratory instability index for one week (every other day) to form a preliminary physiological parameter file. Determine the initial rehabilitation training program based on the OSA severity. Record changes in the patient's physical indicators during the initial rehabilitation training: During a training cycle (one month), continuously monitor indicators such as percentage of hypoxic time, heart rate, respiratory rate, lowest blood oxygen saturation, and nocturnal respiratory instability index, recording the patient's physiological indicator values ​​every other day. The indicators before and after training are compared and used for subsequent multi-parameter joint evaluation to determine the degree of matching between the training plan and the patient's actual pathological characteristics.

[0025] Furthermore, a higher minimum oxygen saturation value indicates milder nocturnal hypoxia and higher sleep quality; AHI indicates the number of apneas or hypoventilations per hour, and a higher value indicates more severe sleep-disordered breathing; the percentage of hypoxic time indicates the proportion of time with nocturnal oxygen saturation below 90% during total sleep time, and a higher value indicates a longer period of nocturnal hypoxia and a higher risk of hypoxia; the nocturnal respiratory instability index indicates a more unstable respiratory rhythm and more severe sleep-disordered breathing.

[0026] Step S102: Based on the physiological indicators of the target user before and during the initial rehabilitation training program, determine the optimized and unoptimized physiological indicators.

[0027] Specifically, since many of the patient's physiological indicators change dynamically during rehabilitation training, by analyzing the trends of these indicators before and after training, when the indicators show an improving trend, it is determined that the current rehabilitation training program has an improving effect on the patient. Furthermore, when an improving trend is identified, the degree of response of the physiological indicators to the current rehabilitation training is determined based on the changes in this trend. If the improvement trend changes from gradual to rapid, even if the overall improvement trend is relatively small, the current rehabilitation training is still considered effective for that physiological indicator.

[0028] Furthermore, as an optional embodiment of the present invention, determining optimized and unoptimized physiological indicators based on the physiological indicators of the target user before and during the initial rehabilitation training program includes: normalizing the physiological indicators of the target user before and during the initial rehabilitation training program to obtain normalized values; performing linear fitting on the normalized values ​​before and during the initial rehabilitation training program to obtain a first trend of change of each physiological indicator before the initial rehabilitation training program and a second trend of change of each physiological indicator during the initial rehabilitation program; determining improvement indicators from the physiological indicators based on the difference between the second and first trends of change of the physiological indicators; determining the improvement enhancement state of the improvement indicators during the initial rehabilitation program using the difference between the second and first trends of change of the improvement indicators and the first trend of change; and determining optimized physiological indicators that meet the physiological state optimization conditions and unoptimized physiological indicators that do not meet the physiological state optimization conditions based on the difference between the second and first trends of change of the improvement indicators, the improvement enhancement state, and the normalized values ​​of the improvement indicators during the initial rehabilitation program.

[0029] Specifically, in this embodiment of the invention, various physiological indicators of the patient are obtained before the initial rehabilitation training program is implemented. These indicators are denoted as follows: ,in, This indicates the patient's baseline physiological parameters before starting the initial rehabilitation training program. This indicates the patient's AHI score before starting the initial rehabilitation training program. This indicates the percentage of time the patient experienced hypoxia before starting the initial rehabilitation training program. This indicates the patient's lowest blood oxygen saturation level before starting the initial rehabilitation training program. This indicates the patient's nocturnal respiratory instability index before starting the initial rehabilitation training program.

[0030] Furthermore, embodiments of the present invention acquire various physiological indicators of the target user during the initial rehabilitation process of the initial rehabilitation training program. , This represents the set of physiological indicators of the target user during the initial rehabilitation process in the initial rehabilitation training program. This indicates the AHI value of the target user during the initial rehabilitation process in the initial rehabilitation training program. This indicates the percentage of time the target user experiences hypoxia during the initial rehabilitation process in the initial rehabilitation training program. This indicates the lowest blood oxygen saturation level of the target user during the initial rehabilitation process in the initial rehabilitation training program. This indicates the nocturnal respiratory instability index of the target user during the initial rehabilitation process of the initial rehabilitation training program.

[0031] Furthermore, embodiments of the present invention use... The normalization process is illustrated using an example: because and Their numerical scoring indicators differ, so to simplify processing, a unified indicator conversion is performed. Then it is converted to ,in, express Normalized values ​​after normalization processing. This indicates that all of the patient's information obtained was... The minimum value among the numerical values. This indicates that all of the patient's information obtained was... The maximum value among the numerical values.

[0032] For others ( ), then convert to .in, express The normalized value of any physiological indicator. express The raw value of any physiological indicator. express The minimum value among all the original values ​​of any physiological indicator. express The maximum value among all the raw values ​​of any physiological indicator.

[0033] Furthermore, for the converted normalized values, the larger the normalized value of this physiological indicator, the higher the quality of the patient's breathing during sleep.

[0034] Similarly, the baseline physiological indicators of patients before initial rehabilitation training... According to the above The same method is used to perform unified indicator conversion, denoted as The embodiments of the present invention will not be described in detail here. Among them, for The normalized set, express Normalized values ​​after normalization express Normalized values ​​after normalization express Normalized values ​​after normalization express Normalized values ​​after normalization.

[0035] Furthermore, the embodiments of the present invention provide a unified index conversion... as well as Perform a linear fit to determine the fitted value. This is represented by the current trends of various physiological indicators for each patient, with the second trend during the initial rehabilitation training process shown as... The first trend of change before initial rehabilitation training is represented as .in, Indicating the initial rehabilitation training process The second trend of change, Indicating the initial rehabilitation training process The second trend of change, Indicating the initial rehabilitation training process The second trend of change, Indicating the initial rehabilitation training process The second trend of change. Indicating the initial rehabilitation training process The first trend of change Indicating the initial rehabilitation training process The first trend of change Indicating the initial rehabilitation training process The first trend of change Indicating the initial rehabilitation training process The first trend of change.

[0036] Furthermore, in this embodiment of the invention, for any given physiological indicator, the current improvement value of that physiological change is determined, that is, the difference between the second trend and the first trend of change of the physiological indicator. ;in, This represents the difference between the second and first trends of change in physiological indicator i. If the physiological trend is improving, then this physiological indicator is identified as an improvement indicator. If so, it indicates that the physiological changes are worsening or the training is ineffective.

[0037] Furthermore, for any improvement indicator Determine the improvement indicator Improvement and enhancement during the initial rehabilitation training process: In the above formula, Indicators of improvement Improvement and enhancement during the initial rehabilitation training process. Indicators of improvement during the initial rehabilitation training process The second trend of change. Indicators of improvement before initial rehabilitation training The first trend of change. Indicators of improvement during the recovery process The degree of improvement. This indicates the degree of improvement of the physiological indicator, which is the patient's initial physiological state before the rehabilitation process, and also indicates the improvement of the patient's physiological state during the initial rehabilitation process. To prevent tiny positive numbers with a denominator of zero, it can take the value 0.001.

[0038] Furthermore, when physiological indicators are in a stable state, their variation tends to decrease after reaching the normal preset threshold. Therefore, the normal preset threshold of the corresponding physiological indicator is introduced as a judgment condition. When the average value of the physiological indicator during the rehabilitation training cycle reaches the normal preset threshold, even if the slope or rate of change of the corresponding physiological indicator is small, it is still determined that the current rehabilitation training has an effective improvement or maintenance effect on the physiological indicator.

[0039] Furthermore, as an optional embodiment of the present invention, the physiological state optimization conditions include: the normalized value of the product of the difference between the second and first trends of the improvement index and the improvement enhancement state is greater than a preset value; or the average value of the normalized values ​​of the improvement index during the initial recovery process is within the normal range of physiological state values.

[0040] Specifically, the preset value in this embodiment of the invention can be 0.65. This embodiment of the invention uses the following formula to represent the physiological state optimization condition: In the above formula, Indicators of improvement Improvement and enhancement during the initial rehabilitation training process. This represents the difference between the second and first trends in the improvement indicator v. The function is used for normalization. This represents the normalized average value of the improvement index v during the initial rehabilitation training process. This represents the minimum value within the normal range of physiological states for the improvement indicator v. This represents the maximum value within the normal range of physiological values ​​for the improvement indicator v.

[0041] Furthermore, in this embodiment of the invention, physiological indicators that do not meet the conditions for optimizing the patient's physiological state are identified as unoptimized physiological indicators and placed in set A, while optimized physiological indicators that meet the conditions for optimizing the physiological state are placed in set B.

[0042] Step S103: Based on the similarity between the first feature vector of the target user and the second feature vector of historical users, determine the first similar user of the target user. The feature vector includes unoptimized physiological indicators and the changing trend during the initial rehabilitation process.

[0043] Specifically, in this embodiment of the invention, since different rehabilitation training programs have different effects on various physiological indicators of patients, it is necessary to screen the following types of samples from historical patient samples: on the one hand, those similar to the target user in terms of improved indicators; on the other hand, those showing a better trend of change in indicators that have not yet improved. There are rehabilitation programs in the rehabilitation training programs of patients with similar physiological changes that have not yet been utilized by the target patient. By analyzing the differences in rehabilitation training programs between the patient samples with similar physiological changes and the target patient, it is possible to determine the effective rehabilitation programs in the current patient's rehabilitation program, as well as the rehabilitation programs required in the current patient's rehabilitation program. Since the optimization status of each physiological indicator and the influence of rehabilitation training programs differ, the effectiveness of each training program can be evaluated by combining the correlation between the current patient's physiological indicators and different rehabilitation programs, as well as the frequency of use of the rehabilitation programs within the patient, in order to determine effective rehabilitation programs.

[0044] Furthermore, as an optional embodiment of the present invention, determining the first similar user of the target user based on the similarity between the first feature vector of the target user and the second feature vector of historical users includes: determining the second similar user of the target user based on the cosine similarity between the first feature vector of the target user and the second feature vector of historical users; and selecting the first similar user from the second similar users based on the difference between the second and first trends of the optimized physiological indicators of the first similar user and the corresponding improvement and enhancement state, and the difference between the second and first trends of the optimized physiological indicators of the target user and the corresponding improvement and enhancement state.

[0045] Specifically, in this embodiment of the invention, the set of unoptimized physiological indicators of the current target user u is first determined. And a set of optimized physiological indicators From all historical users, select those that meet the following criteria: Construct the first feature vector of the current target user u. , which includes the set of its unoptimized physiological indicators Normalized values ​​and first trend of physiological indicators in the data; first eigenvector Mapping is constructed based on a fixed dimensional order of all physiological indicators; for the set For the physiological indicators present in the data, fill in their normalized values ​​and first trend of change in the corresponding dimension positions; for the set... If a corresponding physiological indicator parameter is not found in the database, it will be filled with 0 in the specified dimension. The historical user database will be traversed to calculate the second feature vector of historical patients. and cosine similarity, The normalized values ​​of unoptimized physiological indicators of historical user k and their first trend during the rehabilitation process were included; the top-N patients (e.g., the top 20) with the highest cosine similarity were selected to form a second set of similar users. .

[0046] Furthermore, as an optional embodiment of the present invention, selecting a first similar user from the second similar users based on the difference between the second and first trends of the optimized physiological indicators of the first similar user and the corresponding improvement and enhancement state, and the difference between the second and first trends of the optimized physiological indicators of the target user and the corresponding improvement and enhancement state, includes: selecting a first similar user from the second similar users whose differences between the second and first trends of the optimized physiological indicators and the corresponding improvement and enhancement state are both greater than those of the target user.

[0047] Specifically, the embodiments of the present invention target the second similar user set. For any historical user k in the dataset, determine the set of optimized physiological indicators for that historical user k. The difference between the second and first trends of change corresponding to the v-th optimized physiological indicator. and improve / enhance status Determine the set of optimized physiological indicators for the target user u. The difference between the second and first trends of change corresponding to the v-th optimized physiological indicator. and improve / enhance status .Sure In Are they all greater than In If this condition is not met, then delete the second set of similar users. The corresponding patient is identified as the first similar user. Specifically, the determination of whether the above conditions are met, i.e., the determination of... In Are they all greater than In At that time, if If there is no corresponding improvement indicator for the v-th metric, it is also determined that the condition is not met, and the second similar user set is deleted. The corresponding patient in the middle.

[0048] Step S104: Based on the correlation between the physiological indicators of the first similar user when performing historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user, determine the updated programs for the target user.

[0049] Specifically, in this embodiment of the invention, the training duration of a first similar user during historical rehabilitation training programs is obtained to calculate the correlation between the physiological indicators of the first similar user during historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user. As an optional embodiment of the invention, determining the updated program for the target user based on the correlation between the physiological indicators of the first similar user during historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user includes: obtaining the rehabilitation training programs of the first similar user; obtaining a correlation matrix between the normalized values ​​of the physiological indicators of the first similar user during rehabilitation training programs and the rehabilitation training programs performed by the first similar user; selecting target rehabilitation training programs from the correlation matrix where the correlation between the normalized values ​​of the unoptimized physiological indicators of the target user and the rehabilitation training programs performed by the target user is greater than a preset threshold; determining the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators based on the correlation between the target user's current target rehabilitation training program and the unoptimized and optimized physiological indicators, and the program's experience value; and determining the updated program based on the effectiveness score.

[0050] Specifically, in this embodiment of the invention, the rehabilitation training items of a first similar user are obtained, resulting in a set of training durations for the rehabilitation training items of the first similar user. . This represents the set of training durations for rehabilitation training programs for the first similar users, where... This indicates the training duration of the first rehabilitation training program. This represents the training duration of the nth rehabilitation training program.

[0051] Furthermore, as an optional embodiment of the present invention, the correlation matrix between the normalized values ​​of the physiological indicators of the first similar users when performing rehabilitation training programs and the rehabilitation training programs performed by the first similar users includes: determining a first average value of the training duration of each first similar user performing a rehabilitation training program, and a second average value of the normalized values ​​of the physiological indicators of the first similar users when performing rehabilitation training programs; calculating a first difference between the training duration of each first similar user performing a rehabilitation training program and the first average value, and a second difference between the normalized values ​​of the physiological indicators of the first similar users performing rehabilitation training programs and the second average value; and determining the correlation matrix between each rehabilitation training program and the physiological indicators based on the first difference and the second difference.

[0052] Specifically, the embodiments of the present invention use the following formula to calculate the correlation between each rehabilitation training item and physiological indicators: In the above formula, This represents the correlation between the i-th rehabilitation training program performed by the first similar user and physiological indicator j. This represents the training duration of the i-th rehabilitation training item for the k-th first similar user. This represents the first average training duration of the i-th rehabilitation training program. Let j represent the normalized value of the physiological index j of the kth first similar user. This represents the second average of the normalized values ​​of physiological indicator j. This represents the number of first-like users. `norm` represents the normalization function.

[0053] Based on the above correlation calculation formula, a correlation matrix between rehabilitation training programs and corresponding physiological indicators is constructed: In the above formula, This represents the correlation matrix between rehabilitation training programs and their corresponding physiological indicators. This indicates the correlation between the first rehabilitation training program performed by the first similar user and the first physiological indicator. This indicates the correlation between the first rehabilitation training program performed by the first similar user and the fourth physiological indicator. This represents the correlation between the first rehabilitation training program performed by the first similar user and the nth physiological indicator. This represents the correlation between the nth rehabilitation training program performed by the first similar user and the fourth physiological indicator.

[0054] Furthermore, the preset threshold in this embodiment of the invention can be set to 0.5. This embodiment of the invention targets the unoptimized physiological indicator set of the target user u. Given any physiological indicator j, determine the current correlation matrix. Targeted Rehabilitation Training Programs .in, Indicates the set of unoptimized physiological indicators The correlation between the normalized value of the j-th physiological indicator and the w-th rehabilitation training item.

[0055] Furthermore, as an optional embodiment of the present invention, the method for determining the project experience value includes: the ratio between the first number of first similar users corresponding to the target rehabilitation training project currently being performed by the target user and the second number of all first similar users.

[0056] Specifically, this refers to a set of targeted rehabilitation training programs. Any rehabilitation training program Determined to use the rehabilitation training program The percentage of patients among the first similar users is denoted as the project experience value, calculated using the following formula: In the above formula, Indicates rehabilitation training program Project experience points. This indicates the target rehabilitation training program currently being performed by the target user. The first number of the first similar users. This represents the second number of all first-similar users. The more patients in similar physiological states use the program, the better it improves their physiological state.

[0057] Furthermore, as an optional embodiment of the present invention, the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators is determined based on the correlation between the target user's current target rehabilitation training program and the unoptimized and optimized physiological indicators, and the program's experience value. This includes: The first effective score is determined by multiplying the correlation between the target rehabilitation training program currently being performed by the target user and the unoptimized physiological indicators with the program's experience value; the second effective score is determined by multiplying the correlation between the target rehabilitation training program currently being performed by the target user and the optimized physiological indicators with the program's experience value. The effectiveness score includes the first effective score and the second effective score.

[0058] Specifically, the embodiments of the present invention use the following formula to calculate the first valid score: In the above formula, This indicates the first valid score. This indicates the target rehabilitation training program currently being performed by the target user. Project experience points. This indicates the target rehabilitation training program currently being performed by the target user. Compared with unoptimized physiological indicators The correlation between them.

[0059] Similarly, the optimized set of physiological indicators for target user u The second valid score is calculated using the same method as in the above embodiments.

[0060] Furthermore, as an optional embodiment of the present invention, determining the update items based on the validity score includes: sorting the first valid scores in descending order, selecting the top N items with the first valid scores greater than the validity threshold and the highest ranking as the first update items; sorting the second valid scores in descending order, selecting the top M items with the highest ranking as the second update items, where M is greater than N, and the update items include the first update items and the second update items.

[0061] Specifically, in this embodiment of the invention, N can be 2 and M can be 3. The effective threshold can be determined according to the actual situation, and this embodiment of the invention does not limit it. If the first effective score is not greater than the effective threshold, the target user's current preliminary rehabilitation training plan is maintained. If the top M items with the highest ranking are within the target user's preliminary rehabilitation training plan, the top M items with the highest ranking are retained, and the top N items with the first effective score greater than the effective threshold and the highest ranking are added to the target user's preliminary rehabilitation training plan.

[0062] Step S105: Add updated items to the target user's initial rehabilitation training plan to obtain an updated rehabilitation training plan, and use the updated rehabilitation training plan to conduct rehabilitation training on the target user's upper airway.

[0063] Specifically, if the top M ranked items are already included in the target user's initial rehabilitation training plan, then those top M ranked items are retained. Simultaneously, the top N items with the first valid score greater than the valid threshold and the highest ranking are added to the target user's initial rehabilitation training plan. This results in an updated rehabilitation training plan, which is then used to conduct upper airway rehabilitation training for the target user.

[0064] Furthermore, during each training cycle, the patient's physiological monitoring indicators are recorded, including but not limited to the percentage of time spent in hypoxia, heart rate, and respiratory rate. After each training cycle, the patient's rehabilitation training program for the next cycle is reassessed and determined based on the trends in the monitored physiological indicators. This cycle is continued until the patient recovers or all physiological indicators reach preset normal thresholds, thus achieving personalized adjustment and dynamic optimization of the rehabilitation training program.

[0065] This invention first determines an initial rehabilitation plan based on the severity of OSA. Then, relying on the optimization of the patient's individual physiological indicators and the trend of indicator changes during the rehabilitation process, it identifies unimproved physiological indicators and weaknesses in rehabilitation. This allows for a training plan that is more closely aligned with the patient's unique pathological characteristics and physiological state, ensuring the rehabilitation plan is tailored to the patient's actual condition and improving its suitability. Furthermore, by comparing various physiological indicators before and after the initial rehabilitation training, and during the rehabilitation process, it distinguishes between optimized and unoptimized physiological indicators, identifying weaknesses in the patient's rehabilitation training and avoiding ineffective or redundant training programs. Simultaneously, by combining the similarity between the patient's feature vector and historical user feature vectors, it filters similar users with high matching degrees. Based on rehabilitation data from similar historical cases, it mines the correlation between physiological indicators and training programs, selecting suitable and efficient training programs for target users, taking into account individual differences and clinical patterns, making the development of rehabilitation plans based on evidence. Targeted supplementation with updated training programs effectively compensates for the deficiencies of the initial training plan, making rehabilitation training more targeted and significantly improving training effectiveness. Therefore, the embodiments of the present invention, through personalized optimization and updating of training programs, conduct training in accordance with the patient's actual physiological tolerance and pathological needs. This can not only accurately improve problems such as upper airway collapse, abnormal muscle function, and respiratory regulation disorder, and steadily improve the rehabilitation effect, but also avoid unsuitable training programs that would increase the patient's physical burden, thereby improving rehabilitation efficiency and training safety, and comprehensively improving the overall quality of rehabilitation training.

[0066] Example 2: Corresponding to the personalized upper airway rehabilitation training method based on OSA severity grading provided in the above embodiments, based on the same technical concept, this embodiment of the invention also provides a personalized upper airway rehabilitation training system based on OSA severity grading. This personalized upper airway rehabilitation training system based on OSA severity grading is used to execute the aforementioned personalized upper airway rehabilitation training method based on OSA severity grading. The personalized upper airway rehabilitation training system based on OSA severity grading can vary significantly due to differences in configuration or performance. It may include one or more processors and a memory. The memory stores computer programs that can run on the processor, and the processor executes the programs stored in the memory to achieve the above... Figure 1 The various steps in the Chinese method embodiment. The memory can be temporary or persistent storage. The application stored in the memory may include one or more modules (not shown in the figures), each module may include a series of computer-executable instructions for a personalized upper airway rehabilitation training system based on OSA severity grading.

[0067] Furthermore, the processor can be configured to communicate with memory and execute a series of computer-executable instructions stored in memory on the OSA severity-based personalized upper airway rehabilitation training system. The OSA severity-based personalized upper airway rehabilitation training system may also include one or more power supplies, one or more wired or wireless network interfaces, one or more input / output interfaces, and one or more keyboards.

[0068] Specifically, in this embodiment, the personalized upper airway rehabilitation training system based on OSA severity grading includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, communication interface, and memory communicate with each other via the bus; the memory stores computer programs; and the processor executes the programs stored in the memory to achieve the above... Figure 1 The various steps in the method embodiments are the same as those in the above method embodiments, and have the same beneficial effects. To avoid repetition, the embodiments of the present invention will not be described again here.

[0069] It should be noted that the personalized upper airway rehabilitation training system based on OSA severity grading provided in this embodiment of the invention and the personalized upper airway rehabilitation training method based on OSA severity grading provided in this embodiment of the invention are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned personalized upper airway rehabilitation training method based on OSA severity grading, and has the same or similar beneficial effects. Repeated parts will not be repeated.

Claims

1. A personalized upper airway rehabilitation training method based on OSA severity grading, characterized in that, The personalized upper airway rehabilitation training method based on OSA severity grading includes: Determine an initial rehabilitation training plan based on the severity of OSA in the target user; Based on the target user's physiological indicators before and during the initial rehabilitation training program, optimized and non-optimized physiological indicators are determined. Based on the similarity between the first feature vector of the target user and the second feature vector of historical users, the first similar user of the target user is determined. The feature vector includes unoptimized physiological indicators and the changing trend during the initial recovery process. Based on the correlation between the physiological indicators of the first similar user when performing historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user, the updated programs for the target user are determined. The updated item is added to the initial rehabilitation training plan of the target user to obtain an updated rehabilitation training plan, and the updated rehabilitation training plan is used to conduct rehabilitation training on the upper airway of the target user.

2. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 1, characterized in that, The determination of optimized and non-optimized physiological indicators based on the target user's physiological indicators before and during the initial rehabilitation training program includes: Normalize the physiological indicators of the target user before and during the initial rehabilitation training program to obtain normalized values; Linear fitting was performed on the normalized values ​​before and during the initial rehabilitation training program to obtain the first trend of change of each physiological indicator of the target user before the initial rehabilitation training program and the second trend of change of each physiological indicator during the initial rehabilitation program. Based on the difference between the second trend and the first trend of the physiological indicators, improvement indicators are determined from the physiological indicators; The improvement enhancement status of the improvement index during the initial recovery process is determined by using the difference between the second and first trends of the improvement index and the first trend of the improvement index. Based on the difference between the second and first trends of the improvement indicators, the improvement enhancement state, and the normalized values ​​of the improvement indicators during the initial recovery process, optimized physiological indicators that meet the physiological state optimization conditions and unoptimized physiological indicators that do not meet the physiological state optimization conditions are determined.

3. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 2, characterized in that, The physiological state optimization conditions include: The normalized value of the product of the difference between the second and first trends of the improvement index and the improvement enhancement state is greater than a preset value. Alternatively, the average normalized value of the improvement index during the initial recovery process is within the normal range of physiological conditions.

4. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 1, characterized in that, The step of determining the first similar user of the target user based on the similarity between the first feature vector of the target user and the second feature vector of historical users includes: The second similar user of the target user is determined based on the cosine similarity between the first feature vector of the target user and the second feature vector of historical users; Based on the difference between the second and first trends of the optimized physiological indicators of the first similar user and the corresponding improvement and enhancement state, and the difference between the second and first trends of the optimized physiological indicators of the target user and the corresponding improvement and enhancement state, a first similar user is selected from the second similar users.

5. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 4, characterized in that, The step of selecting the first similar user from the second similar users based on the difference between the second and first trends of the optimized physiological indicators of the first similar users and the corresponding improvement and enhancement status, and the difference between the second and first trends of the optimized physiological indicators of the target user and the corresponding improvement and enhancement status, includes: From the second similar users, select the first similar users whose differences between the second and first trends of optimized physiological indicators and their corresponding improvement and enhancement states are greater than those of the target user.

6. The personalized upper airway rehabilitation training method based on OSA severity grading according to any one of claims 1-4, characterized in that, The step of determining the updated items for the target user based on the correlation between the physiological indicators of the first similar user during historical rehabilitation training programs and the historical rehabilitation training programs performed by the first similar user includes: Obtain the rehabilitation training program of the first similar user; Obtain the normalized values ​​of the physiological indicators of the first similar user when performing the rehabilitation training program and the correlation matrix between the rehabilitation training program performed by the first similar user. Select target rehabilitation training programs from the correlation matrix whose normalized values ​​of the unoptimized physiological indicators of the target user have a correlation greater than a preset threshold with the rehabilitation training programs performed by the target user. Based on the correlation between the target rehabilitation training program currently being performed by the target user and the unoptimized and optimized physiological indicators, and the program's experience value, the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators is determined. The updated items are determined based on the validity score.

7. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 6, characterized in that, The correlation matrix between the normalized values ​​of the physiological indicators of the first similar user when performing the rehabilitation training program and the rehabilitation training program performed by the first similar user includes: Determine a first average of the training duration of each of the first similar users in the rehabilitation training program, and a second average of the normalized values ​​of the physiological indicators of the first similar users when they are in the rehabilitation training program. Calculate the first difference between the training duration of each first similar user in the rehabilitation training program and the first average value, and the second difference between the normalized value of the physiological indicators of the first similar user in the rehabilitation training program and the second average value. Based on the first difference and the second difference, a correlation matrix between each rehabilitation training program and physiological indicators is determined.

8. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 6, characterized in that, The methods for determining the project experience value include: The ratio between the first number of first similar users corresponding to the target rehabilitation training program currently being performed by the target user and the second number of all first similar users.

9. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 6, characterized in that, The step of determining the effectiveness score of the target rehabilitation training program for the unoptimized and optimized physiological indicators based on the correlation between the target user's current target rehabilitation training program and the unoptimized and optimized physiological indicators, and the program's experience value, includes: The first effective score is determined by multiplying the correlation between the target rehabilitation training program currently being performed by the target user and the unoptimized physiological indicators with the program's experience value. The second effective score is determined by multiplying the correlation between the target rehabilitation training program currently being performed by the target user and the optimized physiological indicators with the program experience value. The effectiveness score includes the first effective score and the second effective score.

10. The personalized upper airway rehabilitation training method based on OSA severity grading according to claim 9, characterized in that, The process of determining the updated items based on the validity score includes: Sort the first valid scores in descending order, and select the top N items with the first valid scores greater than the valid threshold and the highest ranking as the first update items; The second valid scores are sorted in descending order, and the top M items with the highest rankings are selected as the second update items, where M is greater than N. The update items include the first update items and the second update items.

Images