A contactless respiratory monitoring method and monitoring system for a neonate

Abstract

The application discloses a contactless breathing monitoring method and system for a newborn. Firstly, an infrared temperature image set is generated by a sliding time window, and a temperature change parameter set is read according to the infrared temperature image set, and then a real-time temperature change period model is generated by fitting. Subsequently, extreme value recognition is performed on the real-time temperature change period model, a breathing frequency parameter and a temperature parameter are calculated, the infrared temperature image set is analyzed, a breathing intensity parameter is calculated, and finally a comprehensive breathing score is generated according to the breathing frequency parameter, the temperature parameter and the breathing intensity parameter, and a breathing state monitoring result is output in combination with a standard judgment model. Based on the characteristics of the fixed posture of the newborn, the breathing of the newborn will cause rhythmic changes in temperature in a certain fixed area, and the above rules can be captured by analyzing the infrared temperature images of the above area, and the analysis of the breathing frequency, body temperature and breathing intensity is realized, and further, the contactless monitoring of the breathing of the newborn is realized.

Description

Technical Field

[0001] This application relates to the field of medical information processing technology, specifically to a non-contact respiratory monitoring method and system for newborns. Background Technology

[0002] Newborns, especially premature infants, low birth weight infants, and critically ill newborns, are the core monitoring subjects in the New Intensive Care Unit (NICU). Respiratory function is a key indicator for assessing their vital signs. Abnormal events such as apnea, respiratory rhythm disorders, and insufficient ventilation, if not identified and intervened in a timely manner, can easily lead to neonatal brain damage or even death. Therefore, accurate, real-time, and continuous monitoring of neonatal respiration has extremely high clinical value.

[0003] Currently, the mainstream neonatal respiratory monitoring equipment used in clinical practice all employ contact-based monitoring methods, mainly including chest and abdominal strap sensors, nasogastric airflow sensors, and electrode-based ECG monitoring derivatives. However, all of these technologies require direct contact with the newborn, and due to the fragility of newborns, the installation and debugging of these devices are quite difficult. Furthermore, most of these devices are designed for adults or young children and cannot meet the monitoring requirements of newborns. Therefore, exploring a contactless respiratory monitoring method for newborns is an urgent research topic in this field. Summary of the Invention

[0004] The main purpose of this application is to provide a non-contact respiratory monitoring method and system for newborns, aiming to solve the deficiency of the lack of non-contact monitoring equipment in the prior art.

[0005] This application achieves the above objectives through the following technical solutions: A non-contact respiratory monitoring method for newborns includes the following steps: Set a sliding time window, and use the sliding time window to capture and generate an infrared temperature image set; Read the temperature change parameter set based on the infrared temperature image set; A real-time temperature change period model is generated by fitting the temperature change parameter set using the least squares method. Extreme value identification is performed on the real-time temperature change cycle model to calculate respiratory rate parameters and temperature parameters; The infrared temperature image set is analyzed to calculate the respiratory intensity parameter; A comprehensive respiratory score is generated based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter. The respiratory status monitoring result is output by combining the standard judgment model and the comprehensive respiratory score.

[0006] Optionally, a sliding time window is set to capture and generate an infrared temperature image set, including the following steps: Set the window length parameter and sliding step parameter for the sliding time window; Acquire a basic set of infrared temperature images; An infrared temperature image set is generated by cropping the basic infrared temperature image set according to the window length parameter and the sliding step parameter.

[0007] Optionally, reading the temperature change parameter set based on the infrared temperature image set includes the following steps: The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone; A sieving threshold is set, and a sieving function is generated based on the sieving threshold and the real-time ambient temperature; wherein the calculation expression of the sieving function is: ,in T 0 represents the real-time ambient temperature, Δ T This indicates the screening threshold, which ranges from 0.5 to 1.5℃. The infrared temperature image set of the breathing zone is sieved according to the sieve function to obtain the infrared temperature image set of the temperature change zone. Temperature identification is performed on the infrared temperature image set of the temperature change region to generate a temperature change parameter set {(T1, t1), (T2, t2), ..., (T... m , t m )}, where m is the infrared temperature image number, T m This represents the temperature variation parameter t of the infrared temperature image numbered m. m Represents the temperature variation parameter T m Sampling time.

[0008] Optionally, the calculation expression for the real-time temperature variation cycle model is as follows: Where T0 represents the real-time ambient temperature. A ω represents amplitude, φ represents angular frequency, φ represents initial phase, and k represents the sliding time window number.

[0009] Optionally, extreme value identification is performed on the real-time temperature change cycle model to calculate respiratory rate parameters and temperature parameters, including the following steps: The real-time temperature variation cycle model is subjected to extreme value identification to obtain a set of temperature extreme values. The calculation expression for the temperature extreme values ​​is as follows: , where n represents the number of the extreme value; The set of extreme temperature values ​​is split using a splitting function to obtain the set of maximum temperature values ​​{(T1, t1), (T2, t2), ..., (T... i , t i )}, wherein the calculation expression of the splitting function is i represents the number of maxima; A respiratory interval parameter set {Δt1, Δt2, ..., Δt} is generated based on the set of temperature maxima. a , Δt i-1 )}, where Δt i-1 =t i -t i-1 , where 'a' represents the number of the breathing interval parameter; The respiratory rate is calculated based on the respiratory interval time parameter set, wherein the expression for calculating the respiratory rate is as follows: ,in , where 'a' represents the number of the breathing interval parameter; Temperature parameters are generated based on the set of temperature maxima, wherein the calculation expression for the temperature parameters is as follows: , where b represents the number of the temperature parameter.

[0010] Optionally, the infrared temperature image set is analyzed to calculate the respiratory intensity parameter, including the following steps: The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone; Obtain a sieving function, and use the sieving function to convert the infrared temperature image set into a temperature-varying region infrared temperature image set; wherein the calculation expression of the sieving function is: ,in T 0 represents the real-time ambient temperature, Δ T This indicates the screening threshold, which ranges from 0.5 to 1.5℃. The area of ​​each image in the infrared temperature image set of the temperature change zone is calculated to generate the area parameters of the temperature change zone; Generate minimum bounding boxes for each image in the infrared temperature image set of the temperature change zone, and generate aspect ratio parameters for the temperature change zone based on the minimum bounding boxes; The area parameter and the aspect ratio parameter of the temperature change zone are normalized respectively to obtain the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone. The respiratory intensity parameter is calculated based on the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone, wherein the calculation expression for the respiratory intensity parameter is as follows: , where α and β represent weight parameters, and α+β=1.

[0011] Optionally, the formula for calculating the area parameter of the temperature variation zone is as follows: Where j represents the image number, m is the infrared temperature image number, and the expression for calculating the normalization parameter of the temperature change region area is: s max and s min Let represent the maximum and minimum areas of the temperature change zone, respectively. The expression for calculating the aspect ratio parameter of the temperature change zone is: The formula for calculating the normalized parameter of the aspect ratio in the temperature-varying region is: , where r0 represents the optimal aspect ratio.

[0012] Optionally, a comprehensive respiratory score is generated based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter. Combining this score with a standard judgment model, the respiratory status monitoring result is output, including the following steps: The respiratory rate and temperature parameters are normalized to generate normalized calculation groups; The comprehensive respiratory score is calculated based on the normalized calculation set; the formula for calculating the comprehensive respiratory score is as follows: , where ω1, ω2 and ω3 all represent weight values, and ω1+ω2+ω3=1; Obtain the standard judgment model, and combine it with the comprehensive respiratory score to output the respiratory status monitoring results. The calculation expression of the standard judgment model is as follows: , where p1 is the decision threshold and is a constant.

[0013] Optionally, the respiratory rate parameter and temperature parameter are normalized to generate a normalized calculation group, including the following steps: The respiratory rate parameter is normalized to generate a normalized respiratory rate value; the normalized expression for the respiratory rate parameter is as follows: ,in Indicates respiratory rate; The temperature parameters are normalized to generate normalized values, where the normalization expression for the temperature parameters is: ,in Indicates temperature parameter; Obtain the respiratory intensity parameter, and generate a normalized calculation group by combining the normalized value of the respiratory rate and the normalized value of the temperature parameter.

[0014] Accordingly, this application also discloses a monitoring system based on the above monitoring method, including: Set a sliding time window, and use the sliding time window to capture and generate an infrared temperature image set; Read the temperature change parameter set based on the infrared temperature image set; A real-time temperature change period model is generated by fitting the temperature change parameter set using the least squares method. Extreme value identification is performed on the real-time temperature change cycle model to calculate respiratory rate parameters and temperature parameters; The infrared temperature image set is analyzed to calculate the respiratory intensity parameter; A comprehensive respiratory score is generated based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter. The respiratory status monitoring result is output by combining the standard judgment model and the comprehensive respiratory score.

[0015] Compared with the prior art, this application has the following beneficial effects: This application first generates an infrared temperature image set by capturing the data through a sliding time window and reads the temperature change parameter set based on the infrared temperature image set. Then, it uses the least squares method to fit the temperature change parameter set to generate a real-time temperature change cycle model. Subsequently, it identifies extreme values ​​in the real-time temperature change cycle model, calculates respiratory rate parameters and temperature parameters, analyzes the infrared temperature image set, calculates respiratory intensity parameters, and finally generates a comprehensive respiratory score based on the respiratory rate parameters, temperature parameters, and respiratory intensity parameters. Combining the standard judgment model and the comprehensive respiratory score, it outputs the respiratory status monitoring results.

[0016] Because newborns lack sufficient motor skills, they are usually placed in a supine position in a crib to ensure smooth breathing. Since the temperature in the neonatal ward is maintained at a relatively constant level year-round, the temperature of the newborn's exhaled air is higher than the ambient temperature. Furthermore, due to the newborn's limited motor skills and relatively fixed posture in the crib, the area where the exhaled air temporarily resides is fixed. In infrared images, when a newborn exhales, a fixed area diagonally above their face will show a temperature increase, and when the newborn inhales, the temperature in this area will decrease. Combined with the rhythm of breathing, the temperature in this area will also show periodic temperature increases and decreases, and these changes can be clearly and accurately captured by infrared images. Furthermore, by analyzing the temperature difference between the aforementioned area and the ambient temperature, as well as the periodic parameters, the intensity of respiration and the newborn's body temperature can be directly reflected, enabling comprehensive monitoring of the respiratory status.

[0017] Compared with existing technologies, this application obtains relevant monitoring parameters by analyzing infrared temperature images. It does not require the installation of a large number of monitoring devices on the newborn, meaning that all monitoring devices will not come into direct contact with the newborn, thus achieving contactless monitoring of the newborn. On the one hand, this can avoid adverse effects on the newborn from the devices and reduce the difficulty of monitoring; on the other hand, it can ensure that all data are real data of the newborn in a natural state, ensuring the accuracy of monitoring.

[0018] Secondly, this application cleverly utilizes the physiological characteristic of newborns having relatively fixed body positions, thereby accurately capturing the rhythmic changes in temperature caused by respiration in a fixed area. The collected signals have high specificity, thus ensuring the accuracy of basic monitoring parameters and effectively improving the accuracy of respiratory monitoring. At the same time, because it is a precise capture of parameters in a specific area, it can also effectively reduce redundant information, which is conducive to improving computational efficiency.

[0019] In the process of analyzing the parameters, this application can not only analyze the respiratory rate through the pattern of temperature change, but also find that the highest temperature in the temperature change zone is related to the temperature of the exhaled gas. That is, the highest temperature in the temperature change zone is directly related to the newborn's body temperature, thereby realizing the monitoring of the newborn's body temperature.

[0020] Meanwhile, the flow rate of exhaled air from a newborn directly affects the size of the diffusion area after the gas is expelled. That is, by analyzing the area and aspect ratio of the temperature change zone, the respiratory intensity can be directly reflected. Therefore, this application can simultaneously obtain the respiratory rate, temperature and respiratory intensity through data analysis, realize multi-dimensional synchronous monitoring, avoid the deviation caused by single-dimensional monitoring, and effectively improve the accuracy of monitoring and analysis. At the same time, by analyzing the above-mentioned dimensional parameters, it can also effectively overcome the limitation of existing technologies that only identify breathing but cannot comprehensively assess the quality of breathing. Finally, by setting a sliding time window, the basic parameters are continuously updated, ensuring that the real-time temperature change cycle model is always dynamically tracking changes, thereby ensuring the accuracy of the model and improving the accuracy of newborn monitoring. Attached Figure Description

[0021] Figure 1 This is a flowchart of a non-contact respiratory monitoring method for newborns disclosed in Embodiment 1 of this application; Figure 2 A simplified diagram of a real-time temperature change cycle model; Figure 3 This is a structural diagram of a monitoring system disclosed in Embodiment 2 of this application; The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] In this invention, unless otherwise explicitly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0025] Implementation Method 1

[0026] Reference Figures 1 to 2 This embodiment, as an optional embodiment of this application, discloses a non-contact respiratory monitoring method for newborns, comprising the following steps: S1. Set a sliding time window and extract and generate an infrared temperature image set through the sliding time window; S11. Set the window length parameter and sliding step parameter of the sliding time window; Set the relevant parameters of the sliding time window as needed, including the window length parameter and the sliding step size parameter. Preferably, the window length parameter is set to 0.5-1.5 min, and the sliding step size parameter is 5-10 s. S12. Obtain the basic infrared temperature image set; S13. Generate an infrared temperature image set by extracting the basic infrared temperature image set according to the window length parameter and the sliding step parameter.

[0027] First, obtain the window length parameter and sliding step parameter set in step S11. Then, trace back from the infrared temperature image with the sampling time as the current time or the closest to the current time. The duration of the back-tracing is the window length parameter. In this way, an infrared temperature image set is extracted from the basic infrared temperature image set and the infrared temperature image set is numbered. After a certain interval, adjust the sliding time window based on the sliding step size parameter and extract the parameters again. It should be noted that the interval can be equal to or greater than the sliding step size parameter, depending on the set parameters.

[0028] Setting a sliding time window can reduce the amount of data required for a single calculation, thereby improving computational efficiency. On the other hand, as time goes on, the value of relevant parameters obtained earlier decreases, and they may also have an adverse effect on the calculation of real-time parameters. Therefore, the sliding time window can clear out the outdated parameters, ensuring that the parameters are always highly matched with the real-time state, thereby improving the accuracy of parameter calculation.

[0029] S2. Read the temperature change parameter set based on the infrared temperature image set; S21. The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone. First, several infrared temperature images are manually selected from the set of infrared temperature images. Each of the selected infrared temperature images is a relevant image of the maximum exhalation state (which corresponds to the peak of exhalation). The images are manually analyzed and a calibration box is generated (the size of the calibration box and its position on the infrared temperature image are set). Since newborns have relatively fixed sleeping positions and limited breathing intensity, rhythmic temperature changes occur within a fixed area. The calibration box generated by analyzing the maximum exhalation state can ensure that the temperature change zone at any time is selected. Therefore, by using the calibration box, the temperature change zone can be quickly captured and selected on each infrared temperature image, generating a set of infrared temperature images of the breathing zone, which effectively improves work efficiency. S22. Set a sieving threshold, and generate a sieving function based on the sieving threshold and the real-time ambient temperature; wherein the calculation expression of the sieving function is: Where T0 represents the real-time ambient temperature, Δ T This indicates the screening threshold, which ranges from 0.5 to 1.5℃.

[0030] Because the exhaled gas is hot, there is always a certain temperature difference between the temperature change zone and the ambient temperature. The screening threshold can be set based on this temperature difference, and the range of the screening threshold can be determined by combining the measured parameters. The preferred value is 0.5-1.5℃.

[0031] Simultaneously, a corresponding screening function can be set based on the screening threshold, wherein the calculation expression of the screening function is: Where T0 represents the real-time ambient temperature, Δ T The threshold value is represented by T(x,y,t); T(x,y,t) represents the temperature of the pixel unit with coordinates (x,y) at time t.

[0032] S23. The infrared temperature image set of the breathing zone is sieved according to the sieve function to obtain the infrared temperature image set of the temperature change zone. Obtain any infrared temperature image from the set of infrared temperature images of the breathing area, and convert it into several pixel units, with each pixel unit having a set of coordinates; By analyzing the pixel unit, the temperature value corresponding to the pixel unit can be obtained. Combined with the sieving function in step S22, the area with substandard temperature can be quickly identified as a non-temperature change area and excluded, thereby generating an infrared temperature image of the temperature change area. By repeating the above steps to screen the infrared temperature image set for each breathing zone, the infrared temperature image set for the temperature change zone can be obtained.

[0033] S24. Perform temperature identification on the infrared temperature image set of the temperature change region to generate a temperature change parameter set {(T1, t1), (T2, t2), ..., (T... m , t m )}, where m is the infrared temperature image number, T m This represents the temperature variation parameter t of the infrared temperature image numbered m. m Represents the temperature variation parameter T m Sampling time.

[0034] Obtain the infrared temperature image set of the temperature-varying region obtained in step S23, and perform temperature analysis on each infrared temperature image of the temperature-varying region. The highest identifiable temperature in each image is taken as the temperature of that infrared temperature image of the temperature-varying region, thereby generating a temperature-varying parameter set {(T1, t1), (T2, t2), ..., (T... m , t m )}, where m is the infrared temperature image number, T m This represents the temperature variation parameter t of the infrared temperature image numbered m. m Represents the temperature variation parameter T m Sampling time; S3. The temperature change parameter set is fitted using the least squares method to generate a real-time temperature change period model; Generate a standard two-dimensional coordinate system, where the horizontal axis represents the sampling time and the vertical axis represents the temperature; Subsequently, based on the temperature change parameter set {(T1, t1), (T2, t2), ..., (T...} obtained in step S24, m , t m In the standard two-dimensional coordinate system, calibration points are set; then, the real-time temperature change cycle model is generated by combining the calibration points with the least squares method. The calculation expression for the real-time temperature change cycle model is as follows: T 0，k A represents the real-time ambient temperature. k ω represents amplitude. k φ represents angular frequency. kThis represents the initial phase, and k represents the sliding time window number; It should be noted that one sliding time window corresponds to one real-time temperature variation period model, and during the process of fitting and generating the real-time temperature variation period model, T 0，k This represents the real-time ambient temperature, which is detected by the corresponding temperature sensor and updated along with the sliding time window. During the fitting process, only the parameters of amplitude, angular frequency, and initial phase need to be fitted.

[0035] S4. Perform extreme value identification on the real-time temperature change cycle model and calculate respiratory rate parameters and temperature parameters; S41. Perform extreme value identification on the real-time temperature change periodic model to obtain a set of temperature extreme values. The calculation expression for the temperature extreme values ​​is as follows: , where n represents the number of the extreme value; First, obtain the real-time temperature variation periodic model obtained in step S3, then calculate its first derivative. Since the solution of the first derivative of the extreme value is 0, the time points corresponding to all extreme values ​​can be quickly calculated, and the temperature values ​​corresponding to each extreme point can be calculated. This leads to the calculation expression for the temperature extreme values, where the calculation expression for the temperature extreme values ​​is: , where n represents the number of the extreme value; S42. The set of extreme temperature values ​​is split using a splitting function to obtain the set of maximum temperature values ​​{(T1, t1), (T2, t2), ..., (T... i , t i )}, wherein the calculation expression of the splitting function is i represents the number of maxima; The extreme points calculated in step S41 include both maxima and minima. Therefore, further filtering is needed to identify the maxima. The calculation expression for the splitting function used to extract the maxima is as follows: ; Similarly, the maximum value can be calculated by further calculating the second derivative; Once the maximum value is obtained, the set of temperature maxima {(T1, t1), (T2, t2), ..., (T...} can be generated. i , t i )}, where i represents the number of maxima; It should be noted that the above-mentioned determination of the maximum value can also be obtained by step-by-step retrieval using a maximum value search formula, the expression of which is: T i >T i-1 And T i >T i+1 ; S43. Generate a respiratory interval time parameter set {Δt1, Δt2, ..., Δt} based on the set of temperature maximum values. a , Δt i-1}, where Δt i-1 =t i -t i-1 , where 'a' represents the number of the breathing interval parameter; Obtain the set of temperature maxima obtained in step S42 {(T1, t1), (T2, t2), ..., (T... i , t i By extracting the time parameter within it, the time difference between any two adjacent maxima can be calculated, where Δt i-1 =t i -t i-1 The time interval between any two adjacent maxima can be obtained using the above time difference calculation formula, thus yielding the respiratory interval time parameter set {Δt1, Δt2, ..., Δt}. a ,…Δt i-1}, where 'a' represents the number of the respiratory interval parameter; S44. Calculate the respiratory rate based on the respiratory interval time parameter set, wherein the expression for calculating the respiratory rate is: ,in , where 'a' represents the number of the breathing interval parameter; Any two adjacent maxima constitute one respiratory cycle. Therefore, by combining the time difference between them, several respiratory rate parameters can be calculated. Finally, the average of these respiratory rate parameters is taken as the final respiratory rate. Therefore, the expression for calculating the respiratory rate is: ,in , where 'a' represents the number of the breathing interval parameter; A newborn's breathing will cause rhythmic changes in temperature within a fixed area. The two adjacent maximum temperature values ​​in the above rhythmic changes constitute a complete respiratory cycle. Therefore, respiratory rate can be calculated by monitoring temperature, eliminating the need for testing with a large number of devices. It cleverly utilizes temperature to achieve non-contact respiratory monitoring and ensures calculation accuracy.

[0036] Calculating respiratory rate by averaging can avoid calculation errors caused by random factors, and it can also unify different frequencies to ensure the orderly progress of subsequent calculations. Furthermore, while calculating the respiratory rate, the respiratory status can also be monitored by analyzing the respiratory rate parameters. The specific methods are as follows: First, obtain the respiratory interval parameter set {Δt1, Δt2, ..., Δt}. a , Δti-1}, and then a time decision domain Δt is set. min -Δt max If several cumulative respiratory interval time parameters exceed the time judgment range or several consecutive respiratory interval time parameters exceed the time judgment range, it is determined that the breathing is abnormal and an alarm is triggered. Under normal breathing conditions, the values ​​of various respiratory interval parameters fluctuate within a certain range. When breathing is rapid, the respiratory interval parameters will decrease; similarly, when breathing is more gradual, the respiratory interval parameters will increase. Therefore, by analyzing the respiratory interval parameters, it is possible to quickly determine whether the breathing state is normal. Secondly, the respiratory interval time parameter can be combined with the time decision domain Δt. min -Δt max The distribution location and changing trend within the respiratory tract enable the determination of respiratory trends; The specific method is as follows, using the time domain Δt min -Δt max The midpoint divides the time judgment domain into an upper and lower half. The distribution of each respiratory interval time parameter is statistically analyzed in the upper and lower half. By combining the points in the upper and lower half and the total number of respiratory interval time parameters, the distribution percentages in the upper and lower half can be calculated. If the distribution percentages in the upper and lower half are basically equal (both close to 50%), the breathing is judged to be normal. If the distribution percentage in the upper half exceeds 80% or other set thresholds, the breathing trend is judged to be rapid. Conversely, the breathing trend is judged to be slow.

[0037] S45. Generate temperature parameters based on the set of temperature maxima, wherein the calculation expression for the temperature parameters is: , where b represents the number of the temperature parameter; Obtain the set of temperature maxima obtained in step S42 {(T1, t1), (T2, t2), ..., (T... i , t i The temperature parameters within the parameter are extracted individually, and the average value of each temperature parameter is output as the final result. The calculation expression for the temperature parameters is as follows: , where b represents the number of the temperature parameter; S5. Analyze the infrared temperature image set and calculate the breathing intensity parameter; S51. The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone. S52. Obtain the sieving function, and convert the infrared temperature image set into a temperature-varying region infrared temperature image set using the sieving function; wherein the calculation expression of the sieving function is: Where T0 represents the real-time ambient temperature, Δ TThis indicates the screening threshold, which ranges from 0.5 to 1.5℃.

[0038] The methods for generating the infrared temperature image set of the temperature-changing region in steps S51 and S52 are exactly the same as the methods in step S2; the infrared temperature image set of the temperature-changing region obtained in step S2 can also be directly called.

[0039] S53. Calculate the area of ​​each image in the infrared temperature image set of the temperature change zone to generate the area parameters of the temperature change zone. Subsequently, each image in the infrared temperature image set of the temperature-changing region is converted into several pixel units. The area parameter of the temperature-changing region can be calculated by counting the number of pixel units. Finally, the average value of each value is calculated as the final area parameter of the temperature-changing region. The expression for calculating the area parameter of the temperature-changing region is as follows: , where j represents the number and m is the infrared temperature image number, that is, an infrared temperature image corresponds to a temperature change area parameter, and finally the average value is taken as the temperature change area parameter corresponding to the sliding time window. S54. Generate minimum bounding boxes for each image in the infrared temperature image set of the temperature change zone, and generate aspect ratio parameters of the temperature change zone based on the minimum bounding boxes. For each image in the infrared temperature image set of the temperature-varying region, a minimum bounding box is generated. The ratio of the length to the width of the minimum bounding box is taken as the aspect ratio parameter of the temperature-varying region corresponding to the infrared temperature image. Finally, the average value is taken as the aspect ratio parameter of the temperature-varying region. The expression for calculating the aspect ratio parameter of the temperature-varying region is as follows: ; S55. Normalize the area parameter of the temperature change zone and the aspect ratio parameter of the temperature change zone respectively to obtain the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone. The area parameter and aspect ratio parameter of the temperature change zone are normalized respectively, and the calculation expression of the normalized area parameter of the temperature change zone is as follows: s max and s min These represent the maximum and minimum areas of the temperature change zone, respectively; The calculation expression for the aspect ratio normalization parameter of the temperature change zone is as follows: , where r0 represents the optimal aspect ratio.

[0040] The above normalization calculation unifies the dimensions of different parameters, which facilitates subsequent calculations. S56. Calculate the respiratory intensity parameter based on the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone, wherein the calculation expression for the respiratory intensity parameter is: , where α and β represent weight parameters, and α+β=1.

[0041] The respiratory intensity parameter is calculated using the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone, wherein the calculation expression for the respiratory intensity parameter is as follows: , where α and β represent weight parameters, and α+β=1; Breathing intensity describes the strength of exhaled airflow. If breathing is strong and powerful, the area of ​​the temperature change zone caused by breathing is larger and the exhaled airflow is faster. Therefore, the temperature change zone is generally elongated and has a larger aspect ratio. Conversely, if breathing is weak, the area of ​​the temperature change zone is smaller and the shape of the temperature change zone tends to be elliptical, and its aspect ratio will also be smaller. Compared with the prior art, this application creatively utilizes temperature change zone images to calculate the area and aspect ratio of the temperature change zone, and then accurately characterizes the breathing intensity through the area and aspect ratio of the temperature change zone, thus achieving rapid monitoring of breathing intensity in a non-contact manner. Secondly, the calculation of respiratory intensity is also based on the temperature change zone image, which eliminates the need to introduce additional basic parameters. This effectively simplifies the calculation process and allows for further exploration of the basic parameters.

[0042] S6. Generate a comprehensive respiratory score based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter. Combine the standard judgment model and the comprehensive respiratory score to output the respiratory status monitoring result.

[0043] S61. Normalize the respiratory rate parameter and temperature parameter to generate a normalized calculation group; S611. Normalize the respiratory rate parameter to generate a normalized respiratory rate value; wherein the normalized expression for the respiratory rate parameter is: ,in Indicates respiratory rate; Obtain the respiratory rate parameter and normalize it to generate a normalized respiratory rate value; the normalization expression for the respiratory rate parameter is as follows: ,in Indicates respiratory rate; S612. Normalize the temperature parameters to generate normalized temperature parameter values, where the normalization expression for the temperature parameters is: ,in This represents the temperature parameter.

[0044] The temperature parameters are normalized to generate normalized values, where the normalization expression for the temperature parameters is: ,in Indicates temperature parameter; S613. Obtain the respiratory intensity parameter, and generate a normalized calculation group by combining the normalized value of the respiratory frequency and the normalized value of the temperature parameter.

[0045] S62. Calculate the comprehensive respiratory score based on the normalized calculation group; the formula for calculating the comprehensive respiratory score is as follows: , where ω1, ω2 and ω3 all represent weight values, and ω1+ω2+ω3=1; The comprehensive respiratory score can be obtained by combining the comprehensive respiratory score and the normalized calculation group. The formula for calculating the comprehensive respiratory score is as follows: , where ω1, ω2 and ω3 all represent weight values, and ω1+ω2+ω3=1; In the calculation of the comprehensive respiratory score, the comprehensive respiratory score is directly related to respiratory rate, temperature parameters and respiratory intensity, thereby realizing a multi-dimensional comprehensive examination of respiratory status, effectively avoiding the shortcomings of single-dimensional monitoring which is greatly affected by occasional factors, and improving the accuracy of monitoring.

[0046] S63. Obtain the standard judgment model, and output the respiratory status monitoring results based on the comprehensive respiratory score. The calculation expression of the standard judgment model is as follows: , where p1 is the decision threshold and is a constant.

[0047] Substituting the comprehensive respiratory score calculated in step S62 into the standard judgment model will output the respiratory status monitoring result corresponding to the current sliding time window; Compared with existing technologies, this application obtains relevant monitoring parameters by analyzing infrared temperature images. It does not require the installation of a large number of devices on the newborn, meaning that all monitoring devices will not come into direct contact with the newborn, thus achieving contactless monitoring of the newborn. On the one hand, it can avoid adverse effects on the newborn from the devices and reduce the difficulty of monitoring; on the other hand, it can ensure that all data are real data of the newborn in a natural state, ensuring the accuracy of monitoring. Secondly, this application cleverly utilizes the physiological characteristic of newborns having relatively fixed body positions, thereby accurately capturing the rhythmic changes in temperature caused by respiration in a fixed area. The collected signals have high specificity, thus ensuring the accuracy of basic monitoring parameters and effectively improving the accuracy of respiratory monitoring. At the same time, since it is an accurate capture of parameters in a specific area, it can also effectively reduce redundant information, which is conducive to improving computational efficiency. Meanwhile, during the analysis of parameters, this application can not only analyze the respiratory rate through the pattern of temperature change, but also find that the highest temperature in the temperature change zone is related to the temperature of the exhaled gas. That is, the highest temperature in the temperature change zone is directly related to the newborn's body temperature, thereby realizing the monitoring of body temperature. Meanwhile, the flow rate of exhaled air from a newborn directly affects the size of the diffusion area after the gas is expelled. That is, by analyzing the area and aspect ratio of the temperature change zone, the respiratory intensity can be directly reflected. Therefore, this application can simultaneously obtain the respiratory rate, temperature and respiratory intensity through data analysis, realize multi-dimensional synchronous monitoring, avoid the deviation caused by single-dimensional monitoring, and effectively improve the accuracy of monitoring and analysis. At the same time, by analyzing the above-mentioned dimensional parameters, it can also effectively overcome the limitation of existing technologies that only identify breathing but cannot comprehensively assess the quality of breathing. Finally, by setting a sliding time window, the basic parameters are continuously updated, thereby ensuring that the real-time temperature change cycle model is always in a dynamic tracking process, thus ensuring the accuracy of the model and improving the accuracy of monitoring.

[0048] Implementation Method 2

[0049] Reference Figure 3 This embodiment, as another optional embodiment of this application, discloses a monitoring system, including a data interception module. The data interception module is used to set a sliding time window and intercept and generate an infrared temperature image set through the sliding time window. The output of the data interception module is communicatively connected to a first analysis module, and the output of the first analysis module is communicatively connected to a data fitting module. The data fitting module is used to fit and generate a real-time temperature change cycle model, and the output of the data fitting module is communicatively connected to a second analysis module. The monitoring system also includes a third analysis module, the input of which is communicatively connected to the data interception module; the outputs of the second and third analysis modules are simultaneously communicatively connected to the monitoring module.

[0050] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A non-contact respiratory monitoring method for newborns, characterized in that, Includes the following steps: Set a sliding time window, and use the sliding time window to capture and generate an infrared temperature image set; Read the temperature change parameter set based on the infrared temperature image set; A real-time temperature change period model is generated by fitting the temperature change parameter set using the least squares method. Extreme value identification is performed on the real-time temperature change cycle model to calculate respiratory rate parameters and temperature parameters; The infrared temperature image set is analyzed to calculate the respiratory intensity parameter; A comprehensive respiratory score is generated based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter. The respiratory status monitoring result is output by combining the standard judgment model and the comprehensive respiratory score.

2. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The process of setting a sliding time window and generating an infrared temperature image set by capturing the data through the sliding time window includes the following steps: Set the window length parameter and sliding step parameter for the sliding time window; Acquire a basic set of infrared temperature images; An infrared temperature image set is generated by cropping the basic infrared temperature image set according to the window length parameter and the sliding step parameter.

3. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The step of reading the temperature change parameter set based on the infrared temperature image set includes the following steps: The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone; A sieving threshold is set, and a sieving function is generated based on the sieving threshold and the real-time ambient temperature; wherein the calculation expression of the sieving function is: Where T0 represents the real-time ambient temperature, and ΔT represents the screening threshold, which ranges from 0.5 to 1.5℃; The infrared temperature image set of the breathing zone is sieved according to the sieve function to obtain the infrared temperature image set of the temperature change zone. Temperature identification is performed on the infrared temperature image set of the temperature change region to generate a temperature change parameter set {(T1, t1), (T2, t2), ..., (T... m , t m )}, where m is the infrared temperature image number, T m This represents the temperature variation parameter t of the infrared temperature image numbered m. m Represents the temperature variation parameter T m Sampling time.

4. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The calculation expression for the real-time temperature change cycle model is as follows: Where T0 represents the real-time ambient temperature. A ω represents amplitude, φ represents angular frequency, φ represents initial phase, and k represents the sliding time window number.

5. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The process of identifying extreme values ​​in the real-time temperature change cycle model and calculating respiratory rate and temperature parameters includes the following steps: The real-time temperature variation cycle model is subjected to extreme value identification to obtain a set of temperature extreme values. The calculation expression for the temperature extreme values ​​is as follows: , where n represents the number of the extreme value; The set of extreme temperature values ​​is split using a splitting function to obtain the set of maximum temperature values ​​{(T1, t1), (T2, t2), ..., (T... i , t i )}, wherein the calculation expression of the splitting function is , where i represents the number of maxima; A respiratory interval parameter set {Δt1, Δt2, ..., Δt} is generated based on the set of temperature maxima. a , Δt i-1 }, where Δt i-1 =t i -t i-1 , where 'a' represents the number of the breathing interval parameter; The respiratory rate is calculated based on the respiratory interval time parameter set, wherein the expression for calculating the respiratory rate is as follows: ,in , where 'a' represents the number of the breathing interval parameter; Temperature parameters are generated based on the set of temperature maxima, wherein the calculation expression for the temperature parameters is as follows: , where b represents the number of the temperature parameter.

6. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The analysis of the infrared temperature image set and the calculation of respiratory intensity parameters include the following steps: The infrared temperature image set is calibrated to generate an infrared temperature image set for the breathing zone; Obtain a sieving function, and use the sieving function to convert the infrared temperature image set into a temperature-varying region infrared temperature image set; wherein the calculation expression of the sieving function is: Where T0 represents the real-time ambient temperature, Δ T This indicates the screening threshold, which ranges from 0.5 to 1.5℃. The area of ​​each image in the infrared temperature image set of the temperature change zone is calculated to generate the area parameters of the temperature change zone; Generate minimum bounding boxes for each image in the infrared temperature image set of the temperature change zone, and generate aspect ratio parameters for the temperature change zone based on the minimum bounding boxes; The area parameter and the aspect ratio parameter of the temperature change zone are normalized respectively to obtain the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone. The respiratory intensity parameter is calculated based on the normalized area parameter and the normalized aspect ratio parameter of the temperature change zone, wherein the calculation expression for the respiratory intensity parameter is as follows: , where α and β represent weight parameters, and α+β=1.

7. The non-contact respiratory monitoring method for newborns according to claim 6, characterized in that, The formula for calculating the area parameter of the temperature change zone is as follows: Where j represents the image number, m is the infrared temperature image number, and the expression for calculating the normalization parameter of the temperature change region area is: s max and s min Let represent the maximum and minimum areas of the temperature change zone, respectively. The expression for calculating the aspect ratio parameter of the temperature change zone is: The formula for calculating the normalized parameter of the aspect ratio in the temperature-varying region is: , where r0 represents the optimal aspect ratio.

8. The non-contact respiratory monitoring method for newborns according to claim 1, characterized in that, The process of generating a comprehensive respiratory score based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter, and then combining the comprehensive respiratory score with a standard judgment model to output the respiratory status monitoring result includes the following steps: The respiratory rate and temperature parameters are normalized to generate normalized calculation groups; The comprehensive respiratory score is calculated based on the normalized calculation set; the formula for calculating the comprehensive respiratory score is as follows: , where ω1, ω2 and ω3 all represent weight values, and ω1+ω2+ω3=1; Obtain the standard judgment model, and combine it with the comprehensive respiratory score to output the respiratory status monitoring results. The calculation expression of the standard judgment model is as follows: , where p1 is the decision threshold and is a constant.

9. The non-contact respiratory monitoring method for newborns according to claim 8, characterized in that, The normalization process for the respiratory rate and temperature parameters to generate normalized calculation sets includes the following steps: The respiratory rate parameter is normalized to generate a normalized respiratory rate value; the normalized expression for the respiratory rate parameter is as follows: ,in Indicates respiratory rate; The temperature parameters are normalized to generate normalized values, where the normalization expression for the temperature parameters is: ,in Indicates temperature parameter; Obtain the respiratory intensity parameter, and generate a normalized calculation group by combining the normalized value of the respiratory rate and the normalized value of the temperature parameter.

10. A monitoring system based on the non-contact respiratory monitoring method for newborns according to any one of claims 1-9, characterized in that, include: The data capture module is used to set a sliding time window and capture and generate an infrared temperature image set through the sliding time window. The first analysis module is used to read the temperature change parameter set based on the infrared temperature image set; The data fitting module is used to fit the temperature change parameter set using the least squares method to generate a real-time temperature change period model. The second analysis module is used to identify extreme values ​​in the real-time temperature change cycle model and calculate respiratory rate parameters and temperature parameters. The third analysis module is used to analyze the infrared temperature image set and calculate the breathing intensity parameters; The monitoring module is used to generate a comprehensive respiratory score based on the respiratory rate parameter, temperature parameter, and respiratory intensity parameter, and output the respiratory status monitoring result by combining the standard judgment model and the comprehensive respiratory score.

Images