Bedside sleep monitoring method and device
Through the non-contact bedside sleep monitoring method, the bedside breathing status signal is analyzed, and the prominent position of sleep is identified and disassembled, the existing contact monitoring methods are solved, and the problems of low comfort and susceptible to receptor dynamic interference are achieved, accurate sleep quality monitoring and targeted intervention are achieved, and health status and quality of life are improved.
Patent Information
- Application Number
- CN202510178645.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-18
- Publication Date
- 2025-05-23
AI Technical Summary
The existing sleep breathing status monitoring methods are mainly contact-type, low comfort and susceptible to dynamic interference, making it difficult to achieve accurate sleep quality monitoring and targeted intervention.
The contactless bedside sleep monitoring method is used to obtain and analyze the respiratory status signals of the bedside, including snoring events and apnea, and combine audio signals and breathing signals to identify and disassemble the prominent position of sleep, thereby achieving health status monitoring and sleep status correction.
Non-contact sleep quality monitoring is realized, reducing physical and dynamic interference, accurately identifying prominent sleep locations, conducting targeted interventions, and improving health status and quality of life.
Smart Images
Figure CN120021944A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent monitoring technology, and in particular to a bedside sleep monitoring method and device. Background Art
[0002] The current sleep breathing status monitoring method is mainly a contact method, which detects the corresponding physiological signals and then monitors sleep breathing through direct contact between the corresponding sensors and the user. The comfort level is low, and the monitoring results are easily disturbed by the user's body movements, which is not conducive to popularization. Therefore, non-contact methods are considered. Snoring, apnea or hypopnea are different manifestations of sleep breathing status. Therefore, the audio signal of the sleeping user can be obtained and analyzed to identify snoring events, and the breathing signal can be obtained and analyzed to identify apnea and hypopnea events. The two are combined to comprehensively monitor their sleep breathing status.
[0003] The causes of poor sleep quality are complex and varied. If they cannot be accurately found, targeted and accurate monitoring and intervention may not be possible, thus affecting the health status and quality of life of patients. Summary of the invention
[0004] To achieve the above objectives, this application provides the following technical solutions:
[0005] According to a first aspect of the present invention, the present invention claims protection for a bedside sleep monitoring method, the method comprising the following steps:
[0006] S1, obtains and decomposes the bedside monitoring signal, and outputs each bedside respiratory status information;
[0007] S2, performing fluctuation data analysis on the bedside respiratory status information and outputting each fluctuation data; outputting the boundary irregularity coefficient of each fluctuation data according to the difference in the amplitude change direction of the discrete points of the boundary of each fluctuation data; obtaining the mutation point in each piece of bedside respiratory status information; screening the protruding sleep position according to the movement of the mutation point in each fluctuation data in the adjacent pieces of bedside respiratory status information, the boundary irregularity coefficient and the boundary amplitude change;
[0008] S3, outputting the overlap index of each sleep protrusion position according to the difference between the shape of each sleep protrusion position and the curling curve; dividing the overlapping sleep protrusion positions and the individual positions of the sleeping person according to the numerical distribution of the overlap index of all sleep protrusion positions; disassembling the overlapping sleep protrusion position into each sub-sleep protrusion position according to the difference in the texture and amplitude change direction within the local window of adjacent discrete points on the boundary of the overlapping sleep protrusion position;
[0009] S4, according to the overlapping index of each sub-sleep protruding position, each sub-sleep protruding position is decomposed into a number of individual positions of the sleeping person; the health status of all the individual positions of the sleeping person is monitored and the sleep status is corrected.
[0010] Furthermore, the method for obtaining the boundary irregularity coefficient of each fluctuation data includes:
[0011] For each fluctuation data, the discrete points on the boundary of the fluctuation data are recorded as boundary discrete points; the gradient direction of each boundary discrete point is obtained; the gradient direction difference between each boundary discrete point and the nearest boundary discrete point is calculated; the average value of the gradient direction difference of all boundary discrete points contained in the fluctuation data is recorded as the boundary irregularity coefficient of each fluctuation data.
[0012] Furthermore, the method for screening the protruding position of sleep includes:
[0013] The optical flow estimation method is used to obtain the fluctuation vector of each mutation point in the bedside respiratory state information; the mean of the fluctuation vectors of all mutation points in each fluctuation data is calculated as the average movement coefficient of each fluctuation data;
[0014] For each fluctuation data, the average movement coefficient of the fluctuation data, the boundary irregularity coefficient and the boundary amplitude change are integrated to output the movement credibility of the fluctuation data;
[0015] Perform threshold analysis on the movement credibility of all fluctuation data in the bedside respiratory state information, and output a disassembly threshold; record the fluctuation data with a movement credibility greater than or equal to the disassembly threshold as a sleep protrusion position.
[0016] Furthermore, the method for obtaining the movement credibility of the fluctuation data includes:
[0017] The moving reliability of the fluctuation data is recorded as A, In the formula, is the average moving coefficient of the fluctuation data; is the maximum value of the average moving coefficient of all fluctuation data; r is the boundary irregularity coefficient of the fluctuation data, and δ is the amplitude variance of all discrete points of the boundary in the fluctuation data.
[0018] Furthermore, outputting the overlap index of each sleep protrusion position according to the difference between the shape of each sleep protrusion position and the curling curve includes:
[0019] For each sleep protrusion position, the number of all discrete points in the sleep protrusion position is obtained, recorded as the area S, and the number of discrete points on the boundary of the sleep protrusion position is obtained, recorded as the perimeter L; according to the numerical comparison of the area and perimeter of the sleep protrusion position, the curling curve similarity factor of the sleep protrusion position is output;
[0020] Obtain the minimum bounding rectangle of each sleep protrusion position, and record the absolute value of the difference between the area of each sleep protrusion position and the number of all discrete points in its minimum bounding rectangle as the bounding rectangle difference of each sleep protrusion position;
[0021] The ratio of the normalized value of the difference in the bounding rectangles of each sleep protrusion position to the similarity factor of the curling curve was taken as the overlap index of each sleep protrusion position.
[0022] Furthermore, the method for obtaining the curling curve similarity factor of the protruding sleep position includes: recording the curling curve similarity factor of the protruding sleep position as f, Wherein, S is the area of the protruding sleeping position; L is the circumference of the protruding sleeping position; π is the pi.
[0023] Furthermore, the division of overlapping sleep protrusion positions and individual positions of sleeping persons according to the numerical distribution of the overlap index of all sleep protrusion positions includes:
[0024] The overlapping indexes of all sleep protruding positions are subjected to threshold analysis to output the overlapping threshold, and the sleep protruding positions with overlapping indexes greater than the overlapping threshold are recorded as overlapping sleep protruding positions; and the sleep protruding positions with overlapping indexes less than or equal to the overlapping threshold are recorded as individual positions of the sleeping person.
[0025] Furthermore, the method of decomposing the overlapping sleep protrusion position into each sub-sleep protrusion position according to the difference in texture and amplitude change direction in the local window of adjacent discrete points on the boundary line of the overlapping sleep protrusion position includes:
[0026] For each overlapping sleep protrusion position, a discrete window of a preset size is constructed with each boundary discrete point of the overlapping sleep protrusion position as the center; texture features of the discrete window are obtained; the gradient direction of each discrete point in the discrete window is obtained; the variance of the gradient direction of all discrete points in the discrete window with each boundary discrete point as the direction difference within the window of each boundary discrete point is recorded;
[0027] Output the concave point credibility of each discrete point on the boundary according to the texture features of each discrete point on the boundary and the direction difference within the window;
[0028] Calculate the sum of the concave point credibility of any two discrete boundary points, recorded as the first sum value; calculate the maximum difference in the gradient direction of all discrete points in the discrete window of the two discrete boundary points; take the difference between the maximum difference and the first sum value as the corresponding concave point credibility between the two discrete boundary points; record the two discrete boundary points with the largest corresponding concave point credibility in the overlapping sleep protrusion position as the two disassembled concave points of the overlapping sleep protrusion position;
[0029] The two disassembly concave points of the overlapping sleep protrusion position are connected to disassemble the overlapping sleep protrusion position into two sub-sleep protrusion positions.
[0030] Furthermore, the outputting of the concave point credibility of each discrete point of the boundary line according to the texture feature of each discrete point of the boundary line and the direction difference within the window includes:
[0031] Let the concave point credibility of the vth discrete point of the boundary be F v , F v =std v ×(|P v -P v-1 |+|M v -M v-1 |); where std v is the direction difference within the window of the vth boundary discrete point; P v , P v-1 Respectively represent the maximum value of the gradient direction of all discrete points in the discrete window centered on the vth and v-1th boundary discrete points; M v 、M v-1 Respectively represent the texture features of the discrete windows centered at the vth and v-1th boundary discrete points;
[0032] The method of decomposing each sub-sleep protrusion position into a plurality of individual positions of the sleeping person according to the overlap index of each sub-sleep protrusion position comprises: repeating the division operation of overlapping the sleep protrusion position and the individual position of the sleeping person for each sub-sleep protrusion position until the overlap index of all the output sub-sleep protrusion positions is less than or equal to the overlap threshold, and outputting all the individual positions of the sleeping person.
[0033] According to the second aspect of the invention, the present invention claims protection for a bedside sleep monitoring device, characterized in that it comprises:
[0034] one or more processors;
[0035] A memory having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, the one or more processors implement the bedside sleep monitoring method.
[0036] The present application relates to the field of intelligent monitoring technology, and in particular to a bedside sleep monitoring method and device, which obtains and decomposes bedside monitoring signals, outputs various pieces of bedside respiratory status information; performs fluctuation data analysis on the bedside respiratory status information and outputs various fluctuation data; outputs the overlap index of each sleep protruding position according to the difference between the shape of each sleep protruding position and the curling curve; divides the overlapping sleep protruding position and the individual position of the sleeping person according to the numerical distribution of the overlap index of all sleep protruding positions; decomposes each sub-sleep protruding position into several individual positions of the sleeping person according to the overlap index of each sub-sleep protruding position; monitors the health status of all individual positions of the sleeping person and corrects the sleep status. The present invention can accurately monitor the sleep quality and correct the sleep status, realize targeted intervention and adjustment, and thus improve the health status and the quality of life of the patient. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 This is a flowchart of a bedside sleep monitoring method claimed in the embodiments of the present application. Detailed implementation manners
[0038] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.
[0039] The terms "first", "second", and "third" in the present application are only used for descriptive purposes, and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", and "third" may explicitly or implicitly include at least one of such features. In the description of the present application, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present application are only used to explain the relative positional relationship, fluctuation conditions, etc. between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes unlisted steps or units, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.
[0040] Referring to "embodiments" herein means that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.
[0041] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Without conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.
[0042] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
[0043] The specific scheme of a bedside sleep monitoring method provided by the present application is described in detail below with reference to the accompanying drawings.
[0044] An embodiment of the present application provides a bedside sleep monitoring method. Specifically, the following bedside sleep monitoring method is provided. Figure 1 , the method comprises the following steps:
[0045] S1, obtains and decomposes the bedside monitoring signal, and outputs each bedside respiratory status information.
[0046] A specific implementation scenario of the embodiment of the present application is an image disassembly scenario of a home bedside image, which is connected to a high-definition camera in the home bedside to obtain a bedside monitoring signal, and the bedside monitoring signal is decomposed line by line, and a 1-second signal is decomposed into h=20 images. Then, each image is subjected to median filtering to eliminate noise in the image, and finally, all images are converted into respiratory state information, and each bedside respiratory state information is output.
[0047] S2, performing fluctuation data analysis on the bedside respiratory status information and outputting each fluctuation data; outputting the boundary irregularity coefficient of each fluctuation data according to the difference in the amplitude change direction of the discrete points of the boundary of each fluctuation data; obtaining the mutation point in each piece of bedside respiratory status information; screening the prominent sleep position according to the movement of the mutation point in each fluctuation data in the adjacent pieces of bedside respiratory status information, the boundary irregularity coefficient and the boundary amplitude change.
[0048] In a real diagnostic environment, factors that make it difficult to identify individual sleeping people include but are not limited to the following: the image feature points are complicated due to background elements, and the skin color of the sleeping person may be similar to the color of the surrounding environment. Therefore, in order to monitor the health status of the sleeping person, it is first necessary to accurately identify the individual sleeping person from the current image.
[0049] Although the skin color of a sleeping person may be similar to the color of the surrounding environment, considering that in the acquired signal, except for the sleeping person who often moves and changes position, the position of the object is fixed and almost never changes. Therefore, the moving foreground image and the fixed background image can be distinguished by the change of discrete points in adjacent strips of the image.
[0050] Specifically, a fluctuation data analysis is performed on the bedside respiratory status information to output each fluctuation data; for each fluctuation data, a discrete point on the boundary of the fluctuation data is recorded as a boundary discrete point; the gradient direction of each boundary discrete point is obtained; the gradient direction difference between each boundary discrete point and the nearest boundary discrete point is calculated; the average value of the gradient direction difference of all boundary discrete points contained in the fluctuation data is recorded as the boundary irregularity coefficient of the fluctuation data.
[0051] Taking the i-th bedside respiratory state information as an example, the i-th bedside respiratory state information is used as input, the fluctuation data in the image is obtained by the fluctuation data extraction algorithm, and the gradient direction of each boundary discrete point is obtained by the Sobel operator. Among them, the difference in the gradient direction of each boundary discrete point and the boundary discrete point closest to it is the absolute value of the difference between the gradient direction of each boundary discrete point and the boundary discrete point closest to it in Euclidean distance. Among them, the fluctuation data analysis and the Sobel operator are well-known technologies, and the specific implementation process will not be repeated.
[0052] If the value of the boundary irregularity coefficient is low, it means that the boundary gradient changes smoothly and the boundary is relatively regular, which is more likely to be background fluctuation data; if the value of the boundary irregularity coefficient is high, it means that the boundary changes irregularly and may be a sleep protrusion position.
[0053] For each piece of bedside respiratory state information, mutation points of the bedside respiratory state information are monitored and each mutation point is output; and the fluctuation vector of each mutation point in the bedside respiratory state information is obtained by using an optical flow estimation method.
[0054] Calculate the mean of the fluctuation vectors of all mutation points in each fluctuation data as the average movement coefficient of each fluctuation data; for each fluctuation data, comprehensively consider the average movement coefficient of the fluctuation data, the boundary irregularity coefficient and the boundary amplitude change, and output the movement credibility of the fluctuation data;
[0055] Specifically, the moving credibility of the fluctuation data is recorded as A, In the formula, is the average moving coefficient of the fluctuation data; is the maximum value of the average moving coefficient of all fluctuation data; r is the boundary irregularity coefficient of the fluctuation data, and δ is the amplitude variance of all discrete points of the boundary in the fluctuation data.
[0056] If the average moving coefficient of the fluctuation data is larger, it means that multiple mutation points have moved in the fluctuation data, and the average moving distance is larger; It reflects the movement of the average movement coefficient of the fluctuation data relative to other fluctuation data in the current bedside respiratory status information. If the ratio is larger, it means that the movement of the fluctuation data is significant in all the fluctuation data; if the boundary irregularity coefficient of the fluctuation data is larger, it means that the boundary changes in the fluctuation data are more irregular, and the more consistent with the characteristics of the fluctuation data of sleeping people; if the amplitude value variance of all boundary discrete points in the fluctuation data is larger, it means that in the fluctuation data, the discrete value difference between each boundary discrete point is larger; the greater the movement credibility of the final output, the greater the possibility that the fluctuation data is a moving sleep protruding position.
[0057] Since the boundary characteristics and mutation point movement between the background fluctuation data and the sleep protrusion position are very different, the foreground and background fluctuation data can be distinguished by the Otsu threshold method.
[0058] Therefore, this embodiment performs threshold analysis on the movement credibility of all fluctuation data in each bedside respiratory state information, and outputs a disassembly threshold; the fluctuation data with a movement credibility greater than or equal to the disassembly threshold is recorded as a sleep protrusion position. Specifically, all movement credibility is used as the input of the Otsu threshold method, and the disassembly threshold is output. The Otsu threshold method is a well-known technology, and the specific implementation process will not be repeated.
[0059] S3, output the overlapping index of each sleep protrusion position according to the difference between the shape of each sleep protrusion position and the curling curve; divide the overlapping sleep protrusion positions and the individual positions of the sleeping person according to the numerical distribution of the overlapping index of all sleep protrusion positions; disassemble the overlapping sleep protrusion positions into each sub-sleep protrusion position according to the difference in texture and amplitude change direction within the local window of adjacent discrete points on the boundary of the overlapping sleep protrusion position.
[0060] Through observation and analysis of the same batch of diagnosed sleepers, it was found that not only the color characteristics of the sleepers were relatively consistent, but also their volume, size and shape were relatively consistent. Therefore, if there is no overlap between the extracted sleep protruding positions, the size of each sleep protruding position in the sleep protruding position should be relatively consistent; if there is overlap between the sleepers, the area of the overlapping sleep protruding position will be significantly larger than the non-overlapping sleep protruding position. At the same time, if there is overlap, the boundary contour of the overlapping sleep protruding position will be more complicated due to the randomness of the position of the sleeper after moving.
[0061] The number of all discrete points in each sleep protrusion position is obtained, recorded as area S, and the number of discrete points on the boundary of the sleep protrusion position is obtained, recorded as perimeter L. According to the numerical comparison of the area and perimeter of the sleep protrusion position, the curling curve similarity factor of the sleep protrusion position is output; the curling curve similarity factor of the sleep protrusion position Wherein, π is the ratio of a circle to a circle, and in this embodiment, π is 3.14.
[0062] The curling curve similarity factor f can reflect the degree of similarity between the protruding position of sleep and the ideal curling curve. If the boundary of the protruding position of sleep is closer to the curling curve, f is closer to the maximum value of 1; if the boundary of the protruding position of sleep is more irregular or complex, f is smaller. It should be noted that the curling curve is the most regular and compact shape. According to the relevant knowledge of the area and perimeter of the curling curve, the ratio of the area and perimeter of the standard curling curve is 4π; for a given area S, the perimeter of the circle is the smallest; if the boundary of the protruding position of sleep is very complex, with depressions or protrusions, then under the same area, its perimeter will respond to increase, and f will become smaller; therefore, if the protruding position of sleep is a standard curling curve, the value of f will take the maximum value of 1; if the shape of the protruding position of sleep is close to the curling curve, the value of f is close to 1; if the shape of the protruding position of sleep is greatly different from the curling curve, or the more serious the boundary distortion, the smaller the value of f.
[0063] Furthermore, the minimum bounding rectangle of each sleep protrusion position is obtained, and the absolute value of the difference between the area of each sleep protrusion position and the number of all discrete points in its minimum bounding rectangle is recorded as the difference in the bounding rectangle of each sleep protrusion position; if the sleep protrusion positions do not overlap, the area difference between the sleep protrusion position and its minimum bounding rectangle is small. The ratio of the normalized value of the bounding rectangle difference of each sleep protrusion position to the curling curve similarity factor is used as the overlap index of each sleep protrusion position. Threshold analysis is performed on the overlap index of all sleep protrusion positions to output the overlap threshold, and the sleep protrusion positions with an overlap index greater than the overlap threshold are recorded as overlapping sleep protrusion positions. The sleep protrusion positions with an overlap index less than or equal to the overlap threshold are recorded as individual positions of the sleeping person. Among them, the Otsu threshold method is used for threshold analysis.
[0064] Among the extracted protruding sleep positions, although the shape of the protruding sleep position of a single sleeping person is not a curled curve, when the sleeping people overlap, the protruding sleep position formed by the overlap is very different from the curled curve, and the boundary is more irregular, and the curled curve similarity factor will be significantly reduced; therefore, if the curled curve similarity factor is smaller, it means that the protruding sleep position is more likely to be an overlapping protruding sleep position. The difference in the bounding rectangle reflects the size of the blank space between the protruding sleep position and its bounding rectangle. If the protruding sleep position does not overlap, there is only a single sleeping person in the protruding sleep position, which is relatively compact, and the difference between the protruding sleep position and its own bounding rectangle is small; if there is an overlap, the overall outline of the protruding sleep position will become more irregular, and the blank space between the protruding sleep position and its own bounding rectangle will increase significantly; therefore, if the difference in the outer rectangle is larger, it means that the overall outline of the protruding sleep position is more irregular, the distribution is more discrete, and the possibility of overlap is greater.
[0065] Since the overlapping protruding sleep positions are generated by multiple sleeping persons in close proximity, and the overlapping parts of the sleeping persons will produce concave positions, and the concave positions correspond to the protruding sleep positions, it is manifested as a large change in the boundary discrete points of the overlapping protruding sleep positions. Therefore, this embodiment uses the concave point disassembly method to disassemble the overlapping protruding sleep positions by finding the concave points in the protruding sleep positions to distinguish the individual sleeping persons. The skin of the sleeping person has texture characteristics, and the texture and skin direction in the adjacent positions are relatively consistent, while the texture characteristics will change at the overlapping parts of different sleeping persons.
[0066] For each overlapping sleep protrusion position, a 5×5 discrete window is constructed with each boundary discrete point of the overlapping sleep protrusion position as the center. The texture feature of the discrete window is obtained; in this implementation, all discrete points in the discrete window are used as input, and the MLBP algorithm is used to obtain the texture feature information in the discrete window.
[0067] Let the concave point credibility of the vth discrete point of the boundary be F v , F v =std v ×(|P v -P v-1 |+|M v -M v-1 |). In the formula, std v is the direction difference within the window of the vth boundary discrete point, expressed as the variance of the gradient direction of all discrete points in the discrete window centered on the vth boundary discrete point; P v , P v-1 Respectively represent the maximum value of the gradient direction of all discrete points in the discrete window centered on the vth and v-1th boundary discrete points; M v 、M v-1 They represent the texture features of the discrete windows centered at the vth and v-1th boundary discrete points respectively.
[0068] The greater the difference in direction within the window, the greater the difference in gradient direction between discrete points within the same window; |P v -P v-1 | reflects the difference between the maximum gradient direction of the current and previous boundary discrete points. The larger the value, the more significant the difference between the gradient direction of the current boundary discrete point and the previous boundary discrete point. |M v -M v-1 | reflects the difference between the texture feature information of the current boundary discrete point and the previous boundary discrete point. The larger the value, the greater the difference in texture features between adjacent discrete points. Therefore, if the concave point credibility F vThe larger it is, the greater the difference between the vth boundary discrete point and the adjacent discrete points, and the greater the possibility that the vth boundary discrete point is a concave point.
[0069] Furthermore, the sum of the concave point credibility of any two discrete boundary points is calculated and recorded as the first sum value; the maximum value difference of the gradient direction of all discrete points in the discrete window of the arbitrary two discrete boundary points is calculated; the smaller the maximum value difference, the more opposite the corresponding gradient direction. The difference between the maximum value difference and the first sum value is taken as the corresponding concave point credibility between the arbitrary two discrete boundary points. The two discrete boundary points with the largest corresponding concave point credibility in the overlapping sleep protrusion position are recorded as the two disassembled concave points of the overlapping sleep protrusion position. The maximum value difference is the absolute value of the difference of the maximum values. The two disassembled concave points of the overlapping sleep protrusion position are connected to disassemble the overlapping sleep protrusion position into two sub-sleep protrusion positions.
[0070] S4, according to the overlapping index of each sub-sleep protruding position, each sub-sleep protruding position is decomposed into a number of individual positions of the sleeping person; the health status of all the individual positions of the sleeping person is monitored and the sleep status is corrected.
[0071] In this embodiment, performing health status monitoring to correct sleep status may include:
[0072] When the health status is difficulty falling asleep, the sleep status is corrected to play white noise to help sleep;
[0073] When the health status is monitored as limited breathing, the sleeping state is corrected to increase ventilation.
[0074] According to the positions of two corresponding concave points, a line is connected to complete the construction of the disassembly line, and the overlapping sleep protrusion position is disassembled. Repeat the division operation of overlapping sleep protrusion position and individual position of sleeping person for each sub-sleep protrusion position until the overlap index of all output sub-sleep protrusion positions is less than or equal to the overlap threshold, and output all individual positions of sleeping persons; specifically, after the overlapping sleep protrusion position is disassembled, its overlap index is calculated again. If the overlap index is still greater than the overlap threshold, the overlapping sleep protrusion position is continued to be disassembled until the overlap index is less than or equal to the overlap threshold.
[0075] After the individual fluctuation data of the sleeping person is divided, the original i-th bedside respiratory status information is image-decomposed according to the size of the fluctuation data and the position in the i-th bedside respiratory status information. After that, the individual images of the sleeping person after division are standardized and used as the input of the convolutional neural network for health status monitoring. The output of the convolutional neural network is the health status of the sleeping person. It should be noted that: the training data set of the convolutional neural network is a public training set of sleeping persons, the optimization algorithm is Adam, and the loss function is the cross entropy loss function; the convolutional layer is 4 layers, followed by 1 fully connected layer, the number of iterations is set to 100 times, and the batch size is 32. The training of the convolutional neural network is a well-known technology and will not be repeated here. A bedside sleep monitoring method is implemented.
[0076] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0077] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present application.
[0078] The specific implementation methods of the invention are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.
Claims
1. A bedside sleep monitoring method, characterized in that: The method comprises the following steps: S1, obtains and decomposes the bedside monitoring signal, and outputs each bedside respiratory status information; S2, performing fluctuation data analysis on the bedside respiratory status information and outputting each fluctuation data; outputting the boundary irregularity coefficient of each fluctuation data according to the difference in the amplitude change direction of the discrete points of the boundary of each fluctuation data; obtaining the mutation point in each piece of bedside respiratory status information; screening the protruding sleep position according to the movement of the mutation point in each fluctuation data in the adjacent pieces of bedside respiratory status information, the boundary irregularity coefficient and the boundary amplitude change; S3, outputting the overlap index of each sleep protrusion position according to the difference between the shape of each sleep protrusion position and the curling curve; dividing the overlapping sleep protrusion positions and the individual positions of the sleeping person according to the numerical distribution of the overlap index of all sleep protrusion positions; disassembling the overlapping sleep protrusion position into each sub-sleep protrusion position according to the difference in the texture and amplitude change direction within the local window of adjacent discrete points on the boundary of the overlapping sleep protrusion position; S4, according to the overlapping index of each sub-sleep protruding position, each sub-sleep protruding position is decomposed into a number of individual positions of the sleeping person; the health status of all the individual positions of the sleeping person is monitored and the sleep status is corrected.
2. A bedside sleep monitoring method as claimed in claim 1, characterized in that: The method for obtaining the boundary irregularity coefficient of each fluctuation data includes: For each fluctuation data, the discrete points on the boundary of the fluctuation data are recorded as boundary discrete points; the gradient direction of each boundary discrete point is obtained; the gradient direction difference between each boundary discrete point and the nearest boundary discrete point is calculated; the average value of the gradient direction difference of all boundary discrete points contained in the fluctuation data is recorded as the boundary irregularity coefficient of each fluctuation data.
3. A bedside sleep monitoring method as claimed in claim 1, characterized in that: The method for screening a prominent sleep position comprises: The optical flow estimation method is used to obtain the fluctuation vector of each mutation point in the bedside respiratory state information; the mean of the fluctuation vectors of all mutation points in each fluctuation data is calculated as the average movement coefficient of each fluctuation data; For each fluctuation data, the average movement coefficient of the fluctuation data, the boundary irregularity coefficient and the boundary amplitude change are integrated to output the movement credibility of the fluctuation data; Perform threshold analysis on the movement credibility of all fluctuation data in the bedside respiratory state information, and output a disassembly threshold; record the fluctuation data with a movement credibility greater than or equal to the disassembly threshold as a sleep protrusion position.
4. A bedside sleep monitoring method as claimed in claim 3, characterized in that: The method for obtaining the movement credibility of the fluctuation data includes: The moving credibility of the fluctuation data is recorded as A, In the formula, is the average moving coefficient of the fluctuation data; is the maximum value of the average moving coefficient of all fluctuation data; r is the boundary irregularity coefficient of the fluctuation data, and δ is the amplitude variance of all discrete points of the boundary in the fluctuation data.
5. A bedside sleep monitoring method as claimed in claim 1, characterized in that: Outputting the overlap index of each sleep protrusion position according to the difference between the shape of each sleep protrusion position and the curling curve includes: For each sleep protrusion position, the number of all discrete points in the sleep protrusion position is obtained, recorded as the area S, and the number of discrete points on the boundary of the sleep protrusion position is obtained, recorded as the perimeter L; according to the numerical comparison of the area and perimeter of the sleep protrusion position, the curling curve similarity factor of the sleep protrusion position is output; Obtain the minimum bounding rectangle of each sleep protrusion position, and record the absolute value of the difference between the area of each sleep protrusion position and the number of all discrete points in its minimum bounding rectangle as the bounding rectangle difference of each sleep protrusion position; The ratio of the normalized value of the difference in the bounding rectangles of each sleep protrusion position to the similarity factor of the curling curve was taken as the overlap index of each sleep protrusion position.
6. A bedside sleep monitoring method as claimed in claim 5, characterized in that: The method for obtaining the curling curve similarity factor of the protruding sleep position comprises: recording the curling curve similarity factor of the protruding sleep position as f, Wherein, S is the area of the protruding sleeping position; L is the circumference of the protruding sleeping position; π is the pi.
7. A bedside sleep monitoring method as claimed in claim 6, characterized in that: The method of dividing the overlapping sleep protruding positions and the individual positions of the sleeping person according to the numerical distribution of the overlap index of all the sleep protruding positions includes: The overlapping indexes of all sleep protruding positions are subjected to threshold analysis to output the overlapping threshold, and the sleep protruding positions with overlapping indexes greater than the overlapping threshold are recorded as overlapping sleep protruding positions; and the sleep protruding positions with overlapping indexes less than or equal to the overlapping threshold are recorded as individual positions of the sleeping person.
8. A bedside sleep monitoring method as claimed in claim 1, characterized in that: The method of decomposing the overlapping sleep protrusion position into each sub-sleep protrusion position according to the difference in texture and amplitude change direction in the local window of adjacent discrete points on the boundary line of the overlapping sleep protrusion position includes: For each overlapping sleep protrusion position, a discrete window of a preset size is constructed with each boundary discrete point of the overlapping sleep protrusion position as the center; texture features of the discrete window are obtained; the gradient direction of each discrete point in the discrete window is obtained; the variance of the gradient direction of all discrete points in the discrete window with each boundary discrete point as the direction difference within the window of each boundary discrete point is recorded; Output the concave point credibility of each discrete point on the boundary according to the texture features of each discrete point on the boundary and the direction difference within the window; Calculate the sum of the concave point credibility of any two discrete boundary points, recorded as the first sum value; calculate the maximum difference in the gradient direction of all discrete points in the discrete window of the two discrete boundary points; take the difference between the maximum difference and the first sum value as the corresponding concave point credibility between the two discrete boundary points; record the two discrete boundary points with the largest corresponding concave point credibility in the overlapping sleep protrusion position as the two disassembled concave points of the overlapping sleep protrusion position; The two disassembly concave points of the overlapping sleep protrusion position are connected to disassemble the overlapping sleep protrusion position into two sub-sleep protrusion positions.
9. A bedside sleep monitoring method as claimed in claim 8, characterized in that: The outputting the concave point credibility of each discrete point of the boundary line according to the texture features of each discrete point of the boundary line and the direction difference within the window includes: Let the concave point credibility of the vth discrete point of the boundary be F v , F v =std v ×(|P v -P v-1 |+|M v -M v-1 |); where std v is the direction difference within the window of the vth boundary discrete point; P v , P v-1 Respectively represent the maximum value of the gradient direction of all discrete points in the discrete window centered on the vth and v-1th boundary discrete points; M v 、M v-1 Respectively represent the texture features of the discrete windows centered at the vth and v-1th boundary discrete points; The method of decomposing each sub-sleep protrusion position into a plurality of individual positions of the sleeping person according to the overlap index of each sub-sleep protrusion position comprises: repeating the division operation of overlapping the sleep protrusion position and the individual position of the sleeping person for each sub-sleep protrusion position until the overlap index of all the output sub-sleep protrusion positions is less than or equal to the overlap threshold, and outputting all the individual positions of the sleeping person.
10. A bedside sleep monitoring device, characterized in that: include: one or more processors; A memory having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement a bedside sleep monitoring method according to any one of claims 1 to 9.