A sleep monitoring method and system based on adenoid hypertrophy in children patients

Abstract

The present application belongs to the technical field of sleep monitoring, and provides a sleep monitoring method and system based on adenoid hypertrophy child patients, which monitors the blood oxygen saturation of the child patients in a historical sleep monitoring period, screens out a mild hypoxia change curve, analyzes the mild hypoxia change curves in multiple historical sleep monitoring periods, obtains a target interval distance stability value, reflects the dynamic stability and trend characteristics of the mild hypoxia state of the child in sleep, helps to identify whether there is long-term fixed amplitude mild hypoxia, provides data support for evaluating the chronic effect of hypoxia on the body, and according to the target interval distance stability value, screens out a key monitoring total interval, so that the monitoring work is focused on the key monitoring total interval, the monitoring efficiency is improved, and potential long-term mild hypoxia problems of the adenoid hypertrophy child patients in the sleep process are found in time.

Description

A method and system for sleep monitoring in children with adenoid hypertrophy Technical Field

[0001] This invention belongs to the field of sleep monitoring technology, specifically a sleep monitoring method and system for children with adenoid hypertrophy. Background Technology

[0002] Adenoid hypertrophy is a common childhood disorder. Excessive adenoid enlargement can obstruct the posterior nasal cavity, affecting normal breathing and leading to a series of sleep problems, with mild hypoxia being a common manifestation. If mild hypoxia is not effectively monitored and intervened for a long period, it may adversely affect a child's growth and development, nervous system function, and cardiovascular system. Therefore, accurate monitoring of the sleep process in children with adenoid hypertrophy, especially monitoring of mild hypoxic cycles, is of significant clinical importance.

[0003] In existing technologies, there is often a lack of precise definition of mild hypoxia, which makes it impossible for the constructed curves to accurately reflect the actual situation of mild hypoxia in children's sleep. In terms of screening the total monitoring interval, existing technologies usually do not fully consider the dynamic stability and trend characteristics of mild hypoxia, and cannot focus on the areas that really need to be focused on, thus reducing monitoring efficiency and making it difficult to discover the potential long-term mild hypoxia problem in children with adenoid hypertrophy during sleep.

[0004] Secondly, regarding the setting of sleep monitoring intervals and time points, existing technologies typically use fixed monitoring intervals, which cannot be dynamically adjusted according to the actual changes in the patient's mild hypoxia. For children with adenoid hypertrophy, their mild hypoxia cycles may change at any time due to various factors (such as sleep posture, respiratory rate, etc.). Fixed monitoring intervals may not be able to capture key changes in time, resulting in monitoring data that does not accurately reflect the patient's condition and delaying treatment.

[0005] Therefore, the present invention provides a sleep monitoring method and system for children with adenoid hypertrophy. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is:

[0008] A sleep monitoring method for children with adenoid hypertrophy includes the following steps:

[0009] During historical sleep monitoring cycles, blood oxygen saturation was monitored in pediatric patients to construct a curve showing changes in mild hypoxia.

[0010] By analyzing the mild hypoxia change curves over multiple historical sleep monitoring cycles, key monitoring intervals were identified.

[0011] The key monitoring interval is divided into a preset key monitoring window. The stability of the multiple key monitoring sub-intervals is analyzed to evaluate whether the division of the multiple key monitoring sub-intervals is reasonable and to determine the optimal key monitoring window.

[0012] Based on the multiple key monitoring sub-intervals determined by the optimal key division window, the stability of the time interval between adjacent key monitoring sub-intervals is analyzed. If it is stable, the sleep monitoring interval value is statically set. If it fluctuates, a switching warning operation is performed to obtain the switching warning value, and the sleep monitoring time point is dynamically set according to the switching warning value.

[0013] As a further aspect of the present invention, the process for constructing the mild hypoxia change curve is as follows:

[0014] The sleep monitoring cycle is divided into several sleep monitoring nodes. The blood oxygen saturation at each sleep monitoring node is obtained, and a blood oxygen saturation change curve is constructed. The local blood oxygen saturation change curve that is maintained in the range of 90% to 95% is selected as the undetermined mild hypoxia change curve.

[0015] Extract the X coordinates of the two endpoints on the curve of mild hypoxia to be determined, perform subtraction, take the absolute value, and output the duration to be determined.

[0016] If the pending duration is less than or equal to the pending duration threshold, it indicates that the child patient has mild hypoxia and is marked as a mild hypoxia change curve.

[0017] A further aspect of this invention is the analysis of mild hypoxia variation curves across multiple historical sleep monitoring cycles, as follows:

[0018] The coordinate points corresponding to the minimum and maximum Y-axis coordinates on each mild hypoxia change curve are selected, and the difference is calculated to output the distance of the mild hypoxia interval.

[0019] Arbitrarily select a mild hypoxia interval distance as the target interval distance, and calculate the ratio of the number of historical sleep monitoring cycles with the target interval distance to the total number of historical sleep monitoring cycles, as the percentage of the number of target intervals;

[0020] The historical sleep monitoring cycles with target interval distances are sorted according to time series. Adjacent historical sleep monitoring cycles are mapped to the order of the historical sleep monitoring cycle sequence and input into the convolutional layer in the convolutional neural network to obtain multiple historical monitoring cycle convolutional layers. These are then input into the Manhattan distance formula to output continuously existing interval difference values.

[0021] As a further aspect of the present invention, the screening process for the key monitoring interval is as follows:

[0022] The ratio of the proportion of target intervals to the difference in consecutive intervals is used to calculate the stable value of the target interval distance. The total interval of mild hypoxia corresponding to the maximum stable value of the target interval distance is selected as the total interval of key monitoring.

[0023] A further aspect of the present invention is as follows: the process of obtaining a preset key focus division window and dividing the total key monitoring interval is as follows:

[0024] Obtain the mild hypoxia change curve corresponding to the key monitoring total interval, and obtain the X-axis coordinates of the two endpoints of the mild hypoxia change curve. Calculate the difference, take the absolute value, and output the total duration corresponding to the key monitoring total interval.

[0025] The average duration of the key monitoring intervals within each historical sleep monitoring cycle is calculated and output as the average duration of key monitoring.

[0026] The average distance of the mild hypoxia zone corresponding to the total key monitoring zone in each historical sleep monitoring cycle is calculated and output as the average key monitoring distance.

[0027] The ratio of the average distance to the average duration of key monitoring is calculated to obtain the unit key division value, which serves as the preset key division window. Based on the unit key division value, the total key monitoring interval is divided to obtain multiple key monitoring sub-intervals.

[0028] A further aspect of this invention is the following: stability analysis is performed on multiple key monitoring sub-intervals, the process of which is as follows:

[0029] After obtaining the ratio of the duration corresponding to each key monitoring sub-interval to the average duration of key monitoring, the average duration of the sub-interval is calculated and output.

[0030] After obtaining the number of blood oxygen saturation parameters monitored in each key monitoring sub-interval and the proportion of the total number of blood oxygen saturation parameters monitored in the total key monitoring interval, the average value of the number of parameters in the sub-interval is calculated and output.

[0031] Based on the order of historical sleep monitoring cycles within the historical sleep monitoring cycle sequence, the average duration of each sub-interval and the average number of parameters in each sub-interval are sorted to construct the sub-interval duration analysis sequence and the sub-interval parameter number analysis sequence.

[0032] The data from the sub-interval duration analysis sequence and the sub-interval parameter quantity analysis sequence are respectively input into the Manhattan distance formula, and the output is the periodic sub-interval duration fluctuation value and the periodic sub-interval parameter quantity fluctuation value.

[0033] As a further aspect of the present invention, the process for determining the optimal focus window is as follows:

[0034] The duration fluctuation value of the periodic sub-interval is summed with the parameter quantity fluctuation value of the periodic sub-interval to output the sub-interval stability analysis value. If the sub-interval stability analysis value is greater than the sub-interval stability analysis threshold, the key division window is re-preset and the evaluation operation is performed until the sub-interval stability analysis value is less than or equal to the sub-interval stability analysis threshold. The preset key division window is then marked as the optimal key division window.

[0035] A further aspect of this invention is the analysis of the time intervals between adjacent key monitoring sub-intervals, as follows:

[0036] Multiple key monitoring sub-intervals within each historical sleep monitoring cycle are sorted from largest to smallest to construct a key monitoring sub-interval sequence. The time interval between adjacent key monitoring sub-intervals within the key monitoring sub-interval sequence is obtained and averaged to output the average interval of the cycle sub-intervals. The standard deviation of the average interval of the cycle sub-intervals corresponding to each historical sleep monitoring cycle is calculated to output the time interval stability assessment value.

[0037] A further aspect of this invention is as follows: if the sleep is stable, the sleep monitoring interval value is statically set; if it fluctuates, a switching warning value is obtained, and the sleep monitoring time point is dynamically set, as follows:

[0038] If the interval stability assessment value is less than or equal to the interval stability assessment threshold, the average value of the interval sub-intervals corresponding to all historical sleep monitoring cycles is averaged and the sleep monitoring interval value is output.

[0039] If the time interval stability assessment value is greater than the time interval stability assessment threshold, a switching warning signal is generated. The change curve corresponding to the key monitoring sub-interval is extracted from the mild hypoxia change curve corresponding to the key monitoring total interval. The sub-interval is divided according to the local change curve corresponding to the adjacent sleep monitoring node to obtain several unit key monitoring sub-curves. The slope of each unit key monitoring sub-curve is obtained.

[0040] The maximum unit key monitoring sub-slope is selected as the switching warning value. The coordinate point corresponding to the blood oxygen saturation currently monitored in the key monitoring sub-interval is extracted and connected with the maximum limit value of the currently monitored key monitoring sub-interval. The X-axis coordinate of the connected coordinate point is used as the sleep monitoring time point.

[0041] A sleep monitoring system for children with adenoid hypertrophy includes the following modules:

[0042] Curve construction module: During historical sleep monitoring cycles, blood oxygen saturation of pediatric patients is monitored to construct a curve of mild hypoxia.

[0043] Analysis and screening module: Analyzes the mild hypoxia change curves in multiple historical sleep monitoring cycles and screens out the key monitoring intervals;

[0044] Window evaluation module: Preset key segmentation window, divide the total key monitoring interval, perform stability analysis on the multiple key monitoring sub-intervals after division, evaluate whether the division of multiple key monitoring sub-intervals is reasonable, and determine the optimal key segmentation window;

[0045] Monitoring settings module: Based on multiple key monitoring sub-intervals determined by the optimal key division window, analyze whether the time interval between adjacent key monitoring sub-intervals is stable. If stable, statically set the sleep monitoring interval value. If fluctuating, execute the switching warning operation, obtain the switching warning value, and dynamically set the sleep monitoring time point according to the switching warning value.

[0046] The beneficial effects of this invention are as follows:

[0047] 1. This invention monitors blood oxygen saturation in pediatric patients during historical sleep monitoring cycles, identifies mild hypoxia change curves, and analyzes these curves across multiple historical sleep monitoring cycles to obtain stable values ​​for target interval distances. This reflects the dynamic stability and trend characteristics of mild hypoxia during sleep in children, helping to identify whether there is long-term, fixed-amplitude mild hypoxia. This provides data support for assessing the chronic effects of hypoxia on the body. Furthermore, based on the stable values ​​for target interval distances, key monitoring intervals are selected, thus not only focusing monitoring efforts on key monitoring intervals and improving monitoring efficiency, but also enabling timely detection of potential long-term mild hypoxia problems in children with adenoid hypertrophy during sleep.

[0048] 2. This invention pre-defines a key monitoring window to divide the total key monitoring interval. Stability analysis is then performed on the multiple key monitoring sub-intervals to assess the rationality of the division and determine the optimal key monitoring window. This not only helps optimize the monitoring strategy and ensures a relatively stable number of blood oxygen saturation parameters within each key monitoring sub-interval, but also helps to screen out key monitoring sub-intervals with relatively stable durations, enabling timely detection of disease trends and providing a strong basis for developing personalized treatment plans. Based on the multiple key monitoring sub-intervals determined by the optimal key monitoring window, the stability of the time interval between adjacent key monitoring sub-intervals is analyzed. If stable, the sleep monitoring interval value is obtained, facilitating long-term tracking and analysis of the patient's mild hypoxia cycle. If fluctuating, a switching warning value is obtained, and the sleep monitoring time point is dynamically set. This allows for flexible adjustment of the monitoring time according to the actual changes in the patient's mild hypoxia, improving the accuracy of capturing key changes in the patient's mild hypoxia and enabling the development of personalized monitoring plans based on each patient's specific situation in different scenarios. Attached Figure Description

[0049] The invention will now be further described with reference to the accompanying drawings.

[0050] Figure 1 is a flowchart of the steps of a sleep monitoring method for children with adenoid hypertrophy according to the present invention.

[0051] Figure 2 is a schematic diagram of a sleep monitoring system for children with adenoid hypertrophy according to the present invention.

[0052] Figure 3 is a flowchart for analyzing whether the time interval between adjacent key monitoring sub-intervals is stable. Detailed Implementation

[0053] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0054] Example 1

[0055] Please refer to Figure 1. An embodiment of the present invention describes a sleep monitoring method for children with adenoid hypertrophy, comprising the following steps:

[0056] Step 1: Monitor blood oxygen saturation of pediatric patients during historical sleep monitoring cycles, construct blood oxygen saturation change curves, and screen out mild hypoxia change curves from the blood oxygen saturation change curves.

[0057] Among them, the historical sleep monitoring cycle represents the total sleep time corresponding to previous sleep processes in children with adenoid hypertrophy;

[0058] In some embodiments, the sleep monitoring cycle is equally divided into several sleep monitoring nodes;

[0059] Specifically, the duration intervals between adjacent sleep monitoring nodes are all equal;

[0060] The blood oxygen saturation at each sleep monitoring point was obtained using a fingertip pulse oximeter, and a blood oxygen saturation change curve was constructed with time as the X-axis and blood oxygen saturation as the Y-axis.

[0061] Those skilled in the art will understand that blood oxygen saturation during mild hypoxia will fluctuate between 90% and 95%.

[0062] The extracted local oxygen saturation change curve, which maintains a blood oxygen saturation range of 90% to 95%, will be used as the curve for undetermined mild hypoxia.

[0063] Extract the X coordinates of the two endpoints on the curve of mild hypoxia to be determined, perform subtraction, take the absolute value, and output the duration to be determined.

[0064] If the duration of pending determination is less than or equal to the threshold of the duration of pending determination, it indicates that the duration of the pending mild hypoxia change curve maintained in the 90% to 95% range is relatively short, indicating that the child patient has mild hypoxia.

[0065] If the duration of the pending determination is greater than the threshold of the duration of the pending determination, it indicates that the duration of the analyzed changes in the pending mild hypoxia is maintained in the range of 90% to 95% for a relatively long time, showing that the pediatric patients do not have mild hypoxia.

[0066] The curve showing the undetermined mild hypoxia change corresponding to the mild hypoxia phenomenon in the child patient will be used as the mild hypoxia change curve.

[0067] Step 2: Analyze the mild hypoxia change curves over multiple historical sleep monitoring cycles to identify the key monitoring intervals;

[0068] It should be noted that each historical sleep monitoring cycle corresponds to a curve showing changes in mild hypoxia;

[0069] In some embodiments, the Y-axis coordinates of all coordinate points on each mild hypoxia change curve are extracted and compared. The coordinate points corresponding to the minimum and maximum Y-axis coordinates are selected and recorded as the total interval of mild hypoxia. The difference between the corresponding maximum and minimum Y-axis coordinates is calculated and the distance of the mild hypoxia interval is output.

[0070] Choose any distance within a mild hypoxia zone as the target zone distance;

[0071] The number of historical sleep monitoring cycles with target interval distances is counted, and the ratio is calculated to the total number of historical sleep monitoring cycles. The output is the percentage of the number of target intervals.

[0072] Multiple historical sleep monitoring cycles are sorted according to time series to obtain a historical sleep monitoring cycle sequence;

[0073] The historical sleep monitoring cycles with target interval distances are sorted according to time series. Adjacent historical sleep monitoring cycles are mapped to the order of the historical sleep monitoring cycle sequence and input into the convolutional layer in the convolutional neural network to obtain multiple historical monitoring cycle convolutional layers.

[0074] The convolutional layers from multiple historical monitoring periods are calculated using the Manhattan distance formula to output the continuously existing interval difference value D. lx ;

[0075] Specifically, the Manhattan distance formula is: Where m represents the total number of convolutional layers in the historical monitoring period. This represents the order of the (q-1)th historical sleep monitoring period within the convolutional layer of the m-th historical monitoring period in the historical sleep monitoring period sequence. This represents the order of the q-th historical sleep monitoring period within the convolutional layer of the m-th historical monitoring period in the historical sleep monitoring period sequence.

[0076] The ratio of the proportion of the target interval to the difference in consecutive intervals is used to calculate the stable value of the target interval distance;

[0077] It is understandable that the stable value of the target interval distance means that by quantifying the correlation between the consistency of the total interval distance of key monitoring and the continuous periodic interval pattern, it reflects the dynamic stability and trend characteristics of mild hypoxia in children's sleep, which helps to identify whether there is long-term fixed amplitude of mild hypoxia and provides data support for assessing the chronic effects of hypoxia on the body.

[0078] Compare the stable values ​​of all target interval distances and select the total mild hypoxia interval corresponding to the largest stable target interval distance as the key monitoring interval.

[0079] Those skilled in the art will understand that the purpose of obtaining the total key monitoring interval is:

[0080] Objective 1: To focus on the stability and continuity of the total monitoring interval by quantitative analysis. The determination of the total monitoring interval helps to identify whether there is a long-term, fixed amplitude of mild hypoxia and to promptly detect potential long-term mild hypoxia in children with adenoid hypertrophy during sleep.

[0081] Objective 2: Determining the key monitoring range makes the monitoring work more focused, avoids unnecessary monitoring and analysis, improves monitoring efficiency, and can improve the accuracy of judging the condition of children patients based on the dynamic stability and trend characteristics of mild hypoxia reflected by the key monitoring range.

[0082] The specific scheme of this embodiment is as follows: During historical sleep monitoring cycles, blood oxygen saturation is monitored in pediatric patients to screen out mild hypoxia change curves. The mild hypoxia change curves in multiple historical sleep monitoring cycles are analyzed to obtain the target interval distance stability value, which reflects the dynamic stability and trend characteristics of mild hypoxia during children's sleep. This helps to identify whether there is long-term, fixed-amplitude mild hypoxia and provides data support for assessing the chronic effects of hypoxia on the body. Moreover, based on the target interval distance stability value, the key monitoring interval is screened out, which not only focuses the monitoring work on the key monitoring interval and improves monitoring efficiency, but also timely detects potential long-term mild hypoxia problems in children with adenoid hypertrophy during sleep.

[0083] Example 2

[0084] Please refer to Figure 1. The sleep monitoring method for children with adenoid hypertrophy according to an embodiment of the present invention further includes the following steps:

[0085] Step 3: Preset the key segmentation window, divide the total key monitoring interval into multiple key monitoring sub-intervals, perform stability analysis on the multiple key monitoring sub-intervals, evaluate whether the division of the multiple key monitoring sub-intervals is reasonable, and determine the optimal key segmentation window;

[0086] In some embodiments, a preset key segmentation window is used, and the process of obtaining multiple key monitoring sub-intervals is as follows:

[0087] For example, obtain the mild hypoxia change curve corresponding to the key monitoring total interval, and obtain the X-axis coordinates of the two endpoints of the mild hypoxia change curve, calculate the difference, take the absolute value, and output the total duration corresponding to the key monitoring total interval.

[0088] The average duration of the key monitoring intervals within each historical sleep monitoring cycle is calculated and output as the average duration of key monitoring.

[0089] The average distance of the mild hypoxia zone corresponding to the total key monitoring zone in each historical sleep monitoring cycle is calculated and output as the average key monitoring distance.

[0090] The ratio of the average key monitoring distance to the average key monitoring duration is calculated, and the resulting unit key classification value is used as the preset key classification window.

[0091] Based on the unit's key value, the total key monitoring interval is divided into multiple key monitoring sub-intervals;

[0092] It should be noted that one key monitoring interval corresponds to one historical sleep monitoring cycle, and after the key monitoring interval is divided according to the unit key division value, multiple key monitoring sub-intervals also exist in one historical sleep monitoring cycle.

[0093] Stability analysis was performed on several key monitoring sub-intervals, as follows:

[0094] After obtaining the ratio of the duration corresponding to each key monitoring sub-interval to the average duration of key monitoring, the average duration of the sub-interval is calculated and output.

[0095] The average duration of each sub-interval is sorted according to the historical sleep monitoring cycle sequence, and a sub-interval duration analysis sequence is constructed.

[0096] The mean duration of adjacent sub-intervals within the sub-interval duration analysis sequence is input into the Manhattan distance formula, and the periodic sub-interval duration fluctuation value D is output. t ;

[0097] Specifically, the Manhattan distance formula is: Where u represents the total number of mean duration values ​​for sub-intervals within the sub-interval duration analysis sequence. This represents the mean duration of the (k-1)th sub-interval within the sub-interval duration analysis sequence. This represents the average duration of the k-th sub-interval within the sub-interval duration analysis sequence;

[0098] Similarly, the number of blood oxygen saturation parameters monitored in each key monitoring sub-interval is obtained, and the ratio is calculated with the total number of blood oxygen saturation parameters monitored in the total key monitoring interval. Then, the average value is calculated and the average number of parameters in the sub-interval is output.

[0099] The mean number of parameters in each sub-interval is sorted according to the order of the historical sleep monitoring cycle within the historical sleep monitoring cycle sequence, and a sub-interval parameter number analysis sequence is constructed.

[0100] The mean number of parameters in adjacent sorted sub-intervals within the sub-interval parameter quantity analysis sequence is input into the Manhattan distance formula, and the periodic sub-interval parameter quantity fluctuation value D is output. sl ;

[0101] Specifically, the Manhattan distance formula is: Where n represents the total number of sub-interval parameter counts, which is the average number of parameters in the sub-interval parameter count analysis sequence. This represents the mean number of parameters in the (r-1)th sub-interval within the sub-interval parameter count analysis sequence. This represents the average number of parameters in the r-th sub-interval within the sub-interval parameter count analysis sequence.

[0102] The stability analysis value of the sub-interval is obtained by summing the fluctuation value of the duration of the periodic sub-interval with the fluctuation value of the number of parameters in the periodic sub-interval.

[0103] It is understandable that the meaning of the sub-interval stability analysis value is: the value obtained by comprehensively analyzing the stability of the key monitoring sub-interval in two dimensions, namely duration and number of blood oxygen saturation parameters. Taking into account the fluctuation of sub-interval duration and the fluctuation of the number of blood oxygen saturation parameters monitored within the sub-interval, it reflects the degree of stability of the key monitoring sub-interval in the periodic monitoring of mild hypoxia in children with adenoid hypertrophy.

[0104] Its function is as follows:

[0105] Function 1: It can measure the fluctuation of the number of blood oxygen saturation parameters within key monitoring sub-intervals, which helps to optimize monitoring strategies, ensure that the number of blood oxygen saturation parameters monitored within each key monitoring sub-interval is relatively stable, and improve the accuracy of providing doctors with patient medical record information;

[0106] Function 2: It can quantify the fluctuation of the duration of key monitoring sub-intervals, which helps to screen out key monitoring sub-intervals with relatively stable durations. This allows doctors to more accurately analyze the periodic characteristics of mild hypoxia in patients based on the data from these stable sub-intervals, promptly detect changes in the condition, and provide a strong basis for developing personalized treatment plans.

[0107] The sub-interval stability analysis value is compared with the sub-interval stability analysis threshold, as follows:

[0108] If the stability analysis value of a sub-interval is greater than the stability analysis threshold of a sub-interval, it indicates that the multiple key monitoring sub-intervals divided according to the unit key division value fluctuate greatly in terms of duration and number of blood oxygen saturation parameters, which is a preset division improvement signal. In this case, the key division window is re-preset and the evaluation operation is carried out until the stability analysis value of the sub-interval is less than or equal to the stability analysis threshold of the sub-interval. The preset key division window is then marked as the best key division window.

[0109] If the stability analysis value of the sub-interval is less than or equal to the stability analysis threshold of the sub-interval, it means that the multiple key monitoring sub-intervals divided according to the unit key division value have small fluctuations in terms of duration and number of blood oxygen saturation parameters, which is displayed as a reasonable signal of preset division, and the preset key division window is marked as the best key division window;

[0110] Step 4: Based on the multiple key monitoring sub-intervals determined by the optimal key division window, analyze whether the time interval between adjacent key monitoring sub-intervals is stable. If it is stable, statically set the sleep monitoring interval value. If it fluctuates, execute the switching warning operation, obtain the switching warning value, and dynamically set the sleep monitoring time point according to the switching warning value.

[0111] In some embodiments, the process of analyzing whether the time interval between adjacent key monitoring sub-intervals is stable is as follows:

[0112] Multiple key monitoring sub-intervals within each historical sleep monitoring cycle are sorted from largest to smallest to construct a sequence of key monitoring sub-intervals;

[0113] For example, if multiple key monitoring sub-intervals within the α historical sleep monitoring cycle are designated as key monitoring sub-interval A, key monitoring sub-interval B, key monitoring sub-interval C, and key monitoring sub-interval D, then the order of the key monitoring sub-interval sequence would be key monitoring sub-interval A, key monitoring sub-interval B, key monitoring sub-interval C, and key monitoring sub-interval D.

[0114] If the key monitoring sub-intervals within the β-historical sleep monitoring cycle are designated as key monitoring sub-interval A, key monitoring sub-interval B, key monitoring sub-interval C, and key monitoring sub-interval D, then the order of the key monitoring sub-interval sequence is key monitoring sub-interval A, key monitoring sub-interval B, key monitoring sub-interval C, and key monitoring sub-interval D.

[0115] Obtain the time interval between adjacent key monitoring sub-intervals within the key monitoring sub-interval sequence, perform mean calculation, and output the mean interval of the periodic sub-intervals;

[0116] The standard deviation of the mean interval of each historical sleep monitoring cycle is calculated, and the time interval stability assessment value is output.

[0117] It is understandable that the time interval stability assessment value means that it reflects the degree of fluctuation in the time interval between adjacent key monitoring sub-intervals in different historical sleep monitoring cycles.

[0118] Its function is to provide data support for setting the sleep monitoring time interval in a static manner, ensure the regularity of the monitoring process, enable doctors to obtain monitoring data at fixed time intervals, and facilitate long-term tracking and analysis of the patient's mild hypoxia cycle;

[0119] If the interval stability assessment value is less than or equal to the interval stability assessment threshold, it indicates that the duration interval distribution between adjacent key monitoring sub-intervals is relatively uniform and stable. The average value of the intervals of all historical sleep monitoring cycles corresponding to the cycle sub-intervals is calculated and the sleep monitoring interval value is output.

[0120] If the time interval stability assessment value is greater than the time interval stability assessment threshold, it indicates that the time interval distribution between adjacent key monitoring sub-intervals is significantly different, and a handover warning signal is generated.

[0121] Based on the switching warning signal, the change curve corresponding to the key monitoring sub-interval is extracted from the mild hypoxia change curve corresponding to the key monitoring total interval, and divided according to the local change curves corresponding to adjacent sleep monitoring nodes to obtain several unit key monitoring sub-curves;

[0122] The slope of the key monitoring sub-curve for each unit is obtained as follows:

[0123] Extract the coordinates of the two endpoints of the key monitoring sub-curve of the unit and input them into the slope calculation formula to output the slope K of the key monitoring sub-curve of the unit. d ;

[0124] Specifically, the formula for calculating the slope is: in, This is represented as the coordinate of one endpoint on the key monitoring sub-curve of the unit. This is represented as the coordinate of another endpoint on the unit's key monitoring sub-curve;

[0125] Compare the slopes of all key monitoring units and select the largest key monitoring unit slope as the switching warning value. Extract the coordinate point corresponding to the blood oxygen saturation currently monitored in the key monitoring sub-interval and connect it with the maximum limit value of the current key monitoring sub-interval according to the switching warning value to obtain the X-axis coordinate of the connected coordinate point, which is used as the sleep monitoring time point.

[0126] It should be noted that the purpose of obtaining sleep monitoring time points is:

[0127] Objective 1: By acquiring switching warning values ​​and dynamically setting sleep monitoring time points, the monitoring time can be flexibly adjusted according to the actual changes in the patient's mild hypoxia, avoiding unnecessary monitoring during periods of no change or slow change, thereby improving monitoring efficiency and rationally allocating monitoring resources.

[0128] Objective 2: By dynamically setting sleep monitoring time points when the intervals are unstable, key changes in mild hypoxia in patients can be captured more promptly, and personalized monitoring plans can be developed based on the specific circumstances of each patient in different scenarios.

[0129] The specific scheme of this embodiment is as follows: A key monitoring window is preset to divide the total key monitoring interval. Stability analysis is then performed on the multiple key monitoring sub-intervals to assess the rationality of the division and determine the optimal key monitoring window. This not only helps optimize the monitoring strategy and ensures that the number of monitored blood oxygen saturation parameters within each key monitoring sub-interval remains relatively stable, but also helps to screen out key monitoring sub-intervals with relatively stable durations, enabling timely detection of disease trends and providing a strong basis for developing personalized treatment plans. Based on the multiple key monitoring sub-intervals determined by the optimal key monitoring window, the stability of the time interval between adjacent key monitoring sub-intervals is analyzed. If stable, the sleep monitoring interval value is obtained, facilitating long-term tracking and analysis of the patient's mild hypoxia cycle. If fluctuating, a switching warning value is obtained, and the sleep monitoring time point is dynamically set. This allows for flexible adjustment of the monitoring time according to the actual changes in the patient's mild hypoxia, improving the accuracy of capturing key changes in the patient's mild hypoxia and enabling the development of personalized monitoring plans based on each patient's specific situation in different scenarios.

[0130] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A sleep monitoring method for children with adenoid hypertrophy, characterized in that: include: During historical sleep monitoring cycles, blood oxygen saturation was monitored in pediatric patients to construct a curve showing changes in mild hypoxia. By analyzing the mild hypoxia change curves over multiple historical sleep monitoring cycles, key monitoring intervals were identified. A preset key monitoring window is used to divide the total key monitoring interval. Stability analysis is then performed on the resulting key monitoring sub-intervals to assess their rationality and determine the optimal key monitoring window. Based on the optimal key monitoring window, the stability of the time interval between adjacent key monitoring sub-intervals is analyzed. If stable, the sleep monitoring interval value is statically set; if fluctuating, a switching warning operation is executed to obtain a switching warning value, and the sleep monitoring time point is dynamically set based on this value. The mild hypoxia change curves within multiple historical sleep monitoring cycles are analyzed as follows: The minimum and maximum Y-axis coordinates on each mild hypoxia change curve are selected, and their differences are calculated to output the mild hypoxia interval distance. An arbitrary mild hypoxia interval distance is selected as the target interval distance, and the total number of historical sleep monitoring cycles with the target interval distance is counted. The ratio is used as the proportion of the target intervals. Historical sleep monitoring cycles with target interval distances are sorted by time series. Adjacent historical sleep monitoring cycles are mapped to the historical sleep monitoring cycle sequence and input into a convolutional layer within a convolutional neural network to obtain multiple historical monitoring cycle convolutional layers. This data is then input into the Manhattan distance formula to output the difference value of consecutive intervals. The selection process for the key monitoring total interval is as follows: The ratio of the proportion of the target intervals to the difference value of consecutive intervals is calculated to obtain a stable value for the target interval distance. The total mild hypoxia interval corresponding to the largest stable target interval distance is selected as the key monitoring total interval. The process of obtaining a preset key segmentation window and dividing the key monitoring total interval is as follows: The mild hypoxia change curve corresponding to the key monitoring total interval is obtained, and the X-axis coordinates of the two endpoints of the mild hypoxia change curve are obtained. The difference is calculated, and the absolute value is taken to output the total duration corresponding to the key monitoring total interval. The average duration of the key monitoring intervals within each historical sleep monitoring cycle is calculated and output as the average duration of key monitoring. The distances of the mild hypoxia intervals corresponding to the total key monitoring intervals within each historical sleep monitoring cycle are averaged to obtain the average key monitoring distance. The ratio between the average key monitoring distance and the average key monitoring duration is calculated to obtain the unit key division value, which serves as the preset key division window. Based on the unit key division value, the total key monitoring intervals are divided to obtain multiple key monitoring sub-intervals.

2. The sleep monitoring method for children with adenoid hypertrophy according to claim 1, characterized in that: The process of constructing the mild hypoxia change curve is as follows: The sleep monitoring cycle is equally divided into several sleep monitoring nodes, the blood oxygen saturation at each sleep monitoring node is obtained, a blood oxygen saturation change curve is constructed, and the local blood oxygen saturation change curve that maintains blood oxygen saturation in the range of 90%~95% is selected as the undetermined mild hypoxia change curve; the X coordinates of the two endpoints of the undetermined mild hypoxia change curve are extracted, the difference is processed, the absolute value is taken, and the undetermined duration is output. If the pending duration is less than or equal to the pending duration threshold, it indicates that the child patient has mild hypoxia and is marked as a mild hypoxia change curve.

3. The sleep monitoring method for children with adenoid hypertrophy according to claim 1, characterized in that: Stability analysis was performed on multiple key monitoring sub-intervals, as follows: The ratio of the duration of each key monitoring sub-interval to the average duration of the key monitoring sub-intervals was obtained, and then averaged to output the average duration of the sub-intervals. The proportion of the number of oxygen saturation parameters monitored in each key monitoring sub-interval to the total number of oxygen saturation parameters monitored in the total key monitoring intervals was obtained, and then averaged to output the average number of parameters in the sub-intervals. Based on the order of historical sleep monitoring cycles within the historical sleep monitoring cycle sequence, the average duration and the average number of parameters for each sub-interval were sorted to construct sub-interval duration analysis sequences and sub-interval parameter quantity analysis sequences. The data from the sub-interval duration analysis sequences and sub-interval parameter quantity analysis sequences were input into the Manhattan distance formula to output the periodic sub-interval duration fluctuation value and the periodic sub-interval parameter quantity fluctuation value.

4. A sleep monitoring method for children with adenoid hypertrophy according to claim 1, characterized in that: The process for determining the optimal focus segmentation window is as follows: Sum the fluctuation value of the duration of the periodic sub-interval with the fluctuation value of the number of parameters in the periodic sub-interval to obtain the sub-interval stability analysis value. If the sub-interval stability analysis value is greater than the sub-interval stability analysis threshold, the focus segmentation window is re-preset and evaluated until the sub-interval stability analysis value is less than or equal to the sub-interval stability analysis threshold. The preset focus segmentation window is then marked as the optimal focus segmentation window.

5. A sleep monitoring method for children with adenoid hypertrophy according to claim 1, characterized in that: The time interval between adjacent key monitoring sub-intervals is analyzed as follows: Multiple key monitoring sub-intervals within each historical sleep monitoring cycle are sorted from largest to smallest to construct a key monitoring sub-interval sequence. The time interval between adjacent key monitoring sub-intervals within the key monitoring sub-interval sequence is obtained and averaged to output the average interval of the cycle sub-intervals. The standard deviation of the average interval of the cycle sub-intervals corresponding to each historical sleep monitoring cycle is calculated to output the time interval stability assessment value.

6. A sleep monitoring method for children with adenoid hypertrophy according to claim 5, characterized in that: If stable, the sleep monitoring interval is statically set. If fluctuating, a switching warning value is obtained, and the sleep monitoring time point is dynamically set. The process is as follows: If the interval stability assessment value is less than or equal to the interval stability assessment threshold, the average value of the intervals of all historical sleep monitoring cycles is averaged and the sleep monitoring interval value is output. If the interval stability assessment value is greater than the interval stability assessment threshold, a switching warning signal is generated. The change curve corresponding to the key monitoring sub-interval is extracted from the mild hypoxia change curve corresponding to the key monitoring total interval, and divided according to the local change curves corresponding to adjacent sleep monitoring nodes to obtain several unit key monitoring sub-curves. The slope of each unit key monitoring sub-curve is obtained. The slope of the largest unit key monitoring sub-curve is selected as the switching warning value. The coordinate point corresponding to the blood oxygen saturation currently monitored in the key monitoring sub-interval is extracted and connected with the maximum limit value of the currently monitored key monitoring sub-interval. The X-axis coordinate of the connected coordinate point is used as the sleep monitoring time point.

7. A sleep monitoring system for children with adenoid hypertrophy, characterized in that: Includes the following modules: The system comprises the following modules: Curve Construction Module: Monitors blood oxygen saturation in pediatric patients during historical sleep monitoring cycles to construct a curve representing mild hypoxia. Analysis and Screening Module: Analyzes the mild hypoxia curves from multiple historical sleep monitoring cycles to screen for key monitoring intervals. Window Evaluation Module: Presets a key segmentation window, divides the key monitoring interval, and performs stability analysis on the resulting key monitoring sub-intervals to evaluate their rationality and determine the optimal key segmentation window. Monitoring Setting Module: Based on the determined optimal key segmentation window and multiple key monitoring sub-intervals, analyzes the stability of the time interval between adjacent key monitoring sub-intervals. If stable, the sleep monitoring interval value is statically set; if fluctuating, a switching warning operation is executed to obtain a switching warning value, and the sleep monitoring time points are dynamically set according to the switching warning value. The analysis of mild hypoxia curves from multiple historical sleep monitoring cycles is as follows: The minimum and maximum Y-axis coordinates on each mild hypoxia curve are selected, and the difference is calculated to output the distance between the mild hypoxia intervals. (The last sentence appears to be a separate, unrelated section and is not translated.) A mild hypoxia interval distance is used as the target interval distance. The ratio of historical sleep monitoring cycles with target interval distances to the total number of historical sleep monitoring cycles is used as the target interval proportion. Historical sleep monitoring cycles with target interval distances are sorted according to time series. Adjacent historical sleep monitoring cycles are mapped to the order of the historical sleep monitoring cycle sequence and input into the convolutional layer in the convolutional neural network to obtain multiple historical monitoring cycle convolutional layers. These are then input into the Manhattan distance formula to output the continuous interval difference value. The selection process for the key monitoring total interval is as follows: The ratio of the target interval proportion to the continuous interval difference value is calculated to obtain the target interval distance stability value. The mild hypoxia total interval corresponding to the largest target interval distance stability value is selected as the key monitoring total interval. The process of obtaining the preset key division window and dividing the key monitoring total interval is as follows: The mild hypoxia change curve corresponding to the key monitoring total interval is obtained, and the X-axis coordinates of the two endpoints of the mild hypoxia change curve are obtained. The difference is calculated, and the absolute value is taken to output the total duration corresponding to the key monitoring total interval. The average duration of the key monitoring intervals within each historical sleep monitoring cycle is calculated and output as the average duration of key monitoring. The distances of the mild hypoxia intervals corresponding to the total key monitoring intervals within each historical sleep monitoring cycle are averaged to obtain the average key monitoring distance. The ratio between the average key monitoring distance and the average key monitoring duration is calculated to obtain the unit key division value, which serves as the preset key division window. Based on the unit key division value, the total key monitoring intervals are divided to obtain multiple key monitoring sub-intervals.

Images