A method for eliminating motion artifacts from heart rate signals for thermal endurance monitoring

Abstract

The present invention relates to the technical field of electronic digital data processing, and particularly relates to a method for eliminating motion artifacts in a heart rate signal for heat tolerance monitoring. During the heat tolerance monitoring process, temperature data, acceleration data, and the PPG heart rate signal of blood vessels of a monitored object are acquired. Different types of each action are determined based on the acceleration data, and the segments of the PPG heart rate signal corresponding to different types of each action are decomposed to determine each IMF component corresponding to each action. According to each IMF component corresponding to each action and the temperature data, the influence of each IMF component corresponding to each action on motion artifacts is determined. Signal reconstruction is performed based on the influence of each IMF component corresponding to each action on motion artifacts to obtain the target PPG heart rate signal with motion artifacts eliminated, avoiding the situation where the heart rate signal is distorted due to motion artifacts, and being beneficial to improving the accuracy of heart rate measurement.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and particularly relates to a method for eliminating motion artifacts in heart rate signals for heat tolerance monitoring. Background Art

[0002] When performing non-contact heart rate signal monitoring, PPG (Photo Plethysmo Graphy) measures heart rate by detecting blood flow changes. As a non-invasive optical technology, PPG emits light of a specific wavelength (such as red, green, or infrared) onto the skin. When the light hits the blood vessels, part of the light energy is absorbed by the blood and part is reflected back to the sensor. The DC component of the PPG signal reflects the average blood volume of the blood and tissues, while the AC component reflects the periodic blood volume changes (heart rate) of the heart.

[0003] Usually, during the process of heat tolerance monitoring, since the monitored object is in a moving state, it will cause changes in the skin contact pressure between the monitoring device and the monitored object. At the same time, movement will also cause rapid changes in blood flow velocity, resulting in obvious motion artifacts in the acquired signal data, thus causing distortion of the measured PPG heart rate signal and affecting the measurement accuracy. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a method for eliminating motion artifacts in heart rate signals for heat tolerance monitoring, and the specific technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for eliminating motion artifacts in heart rate signals for heat tolerance monitoring, including:

[0006] During the heat tolerance monitoring process, obtain the temperature data, acceleration data of the monitored object, and the PPG heart rate signal of the blood vessels;

[0007] Determine each action of different types according to the acceleration data, decompose the segments of the PPG heart rate signal corresponding to each action of different types, and determine each IMF component corresponding to each action;

[0008] Determine the influence of each IMF component corresponding to each action on motion artifacts according to each IMF component corresponding to each action and the temperature data;

[0009] Perform signal reconstruction according to the influence of each IMF component corresponding to each action on motion artifacts to obtain the target PPG heart rate signal with motion artifacts eliminated.

[0010] In one implementation, determining each action of different types according to the acceleration data includes:

[0011] Perform motion decomposition on the acceleration data to obtain motion segments of several actions, and perform normalization processing on each of the motion segments respectively;

[0012] Determine the DTW distance between each of the motion segments after normalization processing, and cluster each of the motion segments according to the DTW distance to determine each action of different types.

[0013] In one implementation, determining the influence of each IMF component corresponding to each action on motion artifacts according to each IMF component corresponding to each action and the temperature data includes:

[0014] Match each IMF component corresponding to each action through DTW to respectively determine several IMF component matching pairs between actions of the same type;

[0015] Determine the action difference situation between actions of the same type according to several of the IMF component matching pairs and the acceleration data;

[0016] Determine the change deviation situation of each IMF component of each action according to the temperature data and the action difference situation, and determine the influence of each IMF component corresponding to each action on motion artifacts according to the change deviation situation of each IMF component of each action.

[0017] In one implementation, determining the action difference situation between actions of the same type according to several of the IMF component matching pairs and the acceleration data includes:

[0018] According to several of the IMF component matching pairs, determine the first difference situation between the actions of each IMF component matching pair under each IMF component;

[0019] Determine the acceleration segment corresponding to each action from the acceleration data, match the acceleration segments corresponding to each action through DTW, respectively determine several acceleration matching pairs between actions of the same type, and determine the second difference situation of the acceleration between the actions of each of the acceleration matching pairs according to several acceleration matching pairs between actions of the same type;

[0020] Wherein, the action difference situation between actions of the same type includes the first difference situation and the second difference situation.

[0021] In one implementation, determining the first difference situation between the actions of each IMF component matching pair under each IMF component according to a plurality of the IMF component matching pairs includes:

[0022] Determine a first action and a second action from various actions respectively, and determine a plurality of candidate IMF component matching pairs corresponding to the first action and the second action according to a plurality of the IMF component matching pairs;

[0023] Determine a first absolute value of the difference in slopes between IMF components in each candidate IMF component matching pair;

[0024] Determine a first amplitude of the IMF component matching the IMF component corresponding to the first action and a second amplitude of the IMF component matching the IMF component corresponding to the second action in each candidate IMF component matching pair respectively, and determine a second absolute value of the difference between the first amplitude and the corresponding second amplitude respectively;

[0025] According to the first absolute value, the second absolute value, and the number of candidate IMF component matching pairs, determine the first difference situation between the first action and the second action under each IMF component in various actions, and return to the step of determining the first action and the second action from various actions respectively until the first difference situation between the actions of each IMF component matching pair under each IMF component is determined.

[0026] In one implementation, determining the second difference situation of accelerations between the actions of each acceleration matching pair according to a plurality of acceleration matching pairs between actions of the same type includes:

[0027] Determine a first action and a second action from various actions respectively, and determine a plurality of candidate acceleration matching pairs corresponding to the first action and the second action according to a plurality of acceleration matching pairs between actions of the same type;

[0028] Determine a third absolute value of the difference in slopes between acceleration segments in each candidate acceleration matching pair;

[0029] Determine a third amplitude of the acceleration segment matching the acceleration segment corresponding to the first action and a fourth amplitude of the acceleration segment matching the acceleration segment corresponding to the second action in each candidate acceleration matching pair respectively, and determine a fourth absolute value of the difference between the third amplitude and the corresponding fourth amplitude respectively;

[0030] Based on the third absolute value, the fourth absolute value, and the number of candidate acceleration matching pairs, determine the second difference situation between the first action and the second action in each type of action, and return the steps of respectively determining the first action and the second action from each type of action until the second difference situation of the accelerations between the actions of each acceleration matching pair is determined.

[0031] In one implementation, the determining the change deviation situation of each IMF component of each action according to the temperature data and the action difference situation, and determining the influence situation of each IMF component corresponding to each action on motion artifacts includes:

[0032] According to the temperature data, determine the temperature DTW matching distance between actions of the same type, and respectively determine the temperature difference situation between actions of the same type according to the natural exponential function and the temperature DTW matching distance;

[0033] Respectively determine the ratio of the first difference situation to the second difference situation, and determine the corresponding variance according to each of the ratios;

[0034] Perform a summation process according to the ratio, the temperature difference situation, and the number of actions in each type of action to obtain a summation result, and obtain the change deviation situation of each IMF component of each action according to the product of the reciprocal of the variance and the summation result;

[0035] According to the change deviation situation of each IMF component of each action, calculate the difference in the change deviation situation between two adjacent actions, and determine the influence situation of each IMF component corresponding to each action on motion artifacts.

[0036] In one implementation, the performing signal reconstruction according to the influence situation of each IMF component corresponding to each action to obtain a target PPG heart rate signal with motion artifacts eliminated includes:

[0037] According to the influence situation of each IMF component corresponding to each action, determine the component weight corresponding to each IMF component;

[0038] Respectively determine the amplitude gap between the corresponding IMF components of two adjacent actions, and respectively determine the reconstruction weight of each IMF component of each action according to the amplitude gap and the component weight corresponding to each IMF component;

[0039] Perform signal reconstruction according to the reconstruction weight of each IMF component of each action to obtain a target PPG heart rate signal with motion artifacts eliminated.

[0040] In one implementation, determining the reconstruction weights of the respective IMF components of each action according to the amplitude difference and the component weights corresponding to each IMF component includes:

[0041] Determining the fluctuation weights of each IMF component respectively according to the natural exponential function and the opposite number of the amplitude difference;

[0042] Determining the reconstruction weights of the respective IMF components of each action respectively according to the product of the fluctuation weight of each IMF component and the component weight.

[0043] In one implementation, performing signal reconstruction according to the reconstruction weights of the respective IMF components of each action to obtain a target PPG heart rate signal with motion artifacts eliminated includes:

[0044] Performing weighted summation on the respective IMF components corresponding to each action according to the reconstruction weights of the respective IMF components of each action to obtain a target PPG heart rate signal with motion artifacts eliminated.

[0045] The present invention has the following beneficial effects:

[0046] By obtaining the temperature data, acceleration data, and PPG heart rate signal of blood vessels of the monitored object during heat tolerance monitoring, determining different types of respective actions according to the acceleration data, decomposing the segments of the PPG heart rate signals corresponding to different types of respective actions, determining the respective IMF components corresponding to each action, determining the influence of the respective IMF components corresponding to each action on motion artifacts according to the respective IMF components corresponding to each action and the temperature data, and performing signal reconstruction according to the influence of the respective IMF components corresponding to each action on motion artifacts to obtain a target PPG heart rate signal with motion artifacts eliminated, the situation that the heart rate signal is distorted due to motion artifacts is avoided, which is beneficial to improving the accuracy of heart rate measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0048] Figure 1 It is a schematic flowchart of the steps of a method for eliminating motion artifacts of a heart rate signal for heat tolerance monitoring provided by an embodiment of the present invention;

[0049] Figure 2 It is a schematic diagram of acceleration data provided by an embodiment of the present invention;

[0050] Figure 3 Schematic diagram of PPG heart rate signal provided by an embodiment of the present invention;

[0051] Figure 4 Schematic diagram of temperature data provided by an embodiment of the present invention. Detailed implementation manners

[0052] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following combines the accompanying drawings and preferred embodiments to detail the specific implementation manners, structures, features and effects of a method for eliminating motion artifacts of heart rate signals for heat tolerance monitoring proposed according to the present invention. In the following description, different "an embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0054] It should be noted that "exemplary" in the embodiments of the present application refers to examples listed for convenience of description, and other embodiments are not limited to the examples listed.

[0055] The following specifically describes the specific solution of a method for eliminating motion artifacts of heart rate signals for heat tolerance monitoring provided by the present invention with reference to the accompanying drawings.

[0056] Please refer to Figure 1 , which shows a flowchart of a method for eliminating motion artifacts of heart rate signals for heat tolerance monitoring provided by an embodiment of the present invention. The method for eliminating motion artifacts of heart rate signals for heat tolerance monitoring may at least include steps S100 - S400:

[0057] S100. During heat tolerance monitoring, obtain the temperature data, acceleration data, and PPG heart rate signal of blood vessels of the monitoring object.

[0058] S200. Determine each action of different types according to the acceleration data, decompose the segments of the PPG heart rate signal corresponding to each action of different types, and determine each IMF component corresponding to each action.

[0059] S300. Determine the influence of each IMF component corresponding to each action on motion artifacts according to each IMF component corresponding to each action and the temperature data.

[0060] S400. According to the influence of each IMF component corresponding to each action on motion artifacts, signal reconstruction is performed to obtain the target PPG heart rate signal with motion artifacts removed.

[0061] In the technical solution of the embodiment of the present application, during the heat tolerance monitoring process, temperature data, acceleration data, and the PPG heart rate signal of blood vessels of the monitored object are acquired. Different types of each action are determined according to the acceleration data, and the segments of the PPG heart rate signal corresponding to different types of each action are decomposed to determine each IMF component corresponding to each action. According to each IMF component corresponding to each action and the temperature data, the influence of each IMF component corresponding to each action on motion artifacts is determined. Signal reconstruction is performed according to the influence of each IMF component corresponding to each action on motion artifacts to obtain the target PPG heart rate signal with motion artifacts removed, avoiding the situation where the heart rate signal is distorted due to motion artifacts, which is beneficial to improving the accuracy of heart rate measurement.

[0062] As Figure 2 , Figure 3 and Figure 4 shown, in one implementation, during the heat tolerance monitoring process, relevant data of the monitored object can be acquired by using a monitoring device. For example, by placing a portable ear-hanging device around the ear of the monitored object, the PPG heart rate signal and temperature data of the blood vessels of the monitored object are collected. As Figure 3 shown, Figure 3 is a schematic diagram of the PPG heart rate signal. After magnifying a segment of the PPG heart rate signal 31, one cardiac cycle 32 can be seen (the ordinate of the PPG heart rate signal is dimensionless and is a relative value), which reflects the change ratio of the light intensity received by the optoelectronic sensor). As Figure 4 shown, Figure 4 is a schematic diagram of the temperature data. It can be seen from the temperature data that the temperature (body temperature) changes corresponding to different times, and the change of the set point can be seen, as well as information such as which time period corresponds to shivering, vasoconstriction, sweating, vasodilation, etc.; at the same time, as Figure 2 shown, Figure 2 is a schematic diagram of the acceleration data. The acceleration data of the monitored object can be acquired by using a motion module with an acceleration sensor and a gyroscope. Different monitored objects can have different acceleration data, that is, each acceleration corresponding to different times. It should be noted that the monitoring time length is adjusted based on the actual situation, and the monitoring device can use existing devices and modules, which will not be described in detail.

[0063] It should be noted that during the thermal endurance monitoring, when obtaining the PPG heart rate signal of blood vessels for heart rate signal monitoring, due to the certain movement of the monitoring object, the movement will cause certain motion artifacts in the obtained PPG heart rate signal. When removing the artifacts, in fact, the actions of the monitoring object can be decomposed, and the PPG heart rate signals corresponding to the same actions can be analyzed. Although the motion artifacts of the PPG heart rate signals of the same actions may be affected by temperature and the frequency of heart rate beats during acquisition, the PPG heart rate signals of the same actions have certain rules. Therefore, the present application analyzes by means of EMD decomposition to determine the influence of different actions on motion artifacts.

[0064] In the embodiment of the present application, based on the principle that different actions will result in different interferences of motion artifacts, quantitative analysis is carried out, and action classification is performed based on acceleration data.

[0065] In one implementation manner, in step S200:

[0066] First, the acceleration data can be decomposed by an existing method to obtain motion segments of several actions. Each motion segment has a certain number of accelerations (values), and then each motion segment is normalized respectively to avoid the interference caused by the motion intensity.

[0067] Secondly, the DTW distance between each normalized motion segment is determined. DTW (Dynamic Time Warping) is an algorithm for measuring the similarity of two time series, and then the DTW distance is used as the distance during clustering. Each motion segment is clustered according to the DTW distance, so as to determine different types and each action in different types.

[0068] Then, by means of the EMD (Empirical Mode Decomposition) decomposition method, the segments of the PPG heart rate signals corresponding to each action of different types are decomposed to determine each IMF (Intrinsic Mode Function) component corresponding to each action. Each IMF component is equivalent to a part of the signal of the decomposed PPG heart rate signal.

[0069] In the embodiments of the present application, the acquired PPG heart rate signal itself is composed of motion artifacts, noise, and the heart rate part. They are superimposed on each other and not easily distinguishable. Therefore, the EMD decomposition method is used to operate on its components. At the same time, by analyzing the different IMF components after decomposition, the influence of temperature can be reduced to a certain extent (temperature affects the trend of IMF components). Among them, when analyzing the IMF components subsequently, the IMF components in the high-frequency region can be removed because the heart rate is generally distributed in the middle-frequency region, and the low-frequency region is generally caused by the changes in the respiration and body temperature of the monitored object (changes in trend). Therefore, when analyzing the IMF components subsequently, it refers to analyzing the middle-frequency region and the part of the low-frequency region of the IMF components.

[0070] It should be noted that the above division of the same type of actions, the analysis of the generation of motion artifacts is related to the motion actions. Under the same action, there is a certain relationship between the acquired acceleration data and the PPG heart rate signal. The stability and noise influence of the PPG heart rate signal under the same acceleration can be analyzed. In addition, since the influence of motion artifacts is also related to the temperature of the monitored object, the intensity of its motion artifacts will increase after the temperature rises. Therefore, the influence of different temperatures under the same action needs to be considered when performing subsequent quantitative analysis.

[0071] In one embodiment, step S300 includes steps S301 - S303:

[0072] S301. Match each IMF component corresponding to each action through DTW, and respectively determine several IMF component matching pairs between actions of the same type.

[0073] Optionally, match each IMF component corresponding to each action through DTW, and respectively determine several IMF component matching pairs between actions of the same type; among them, each IMF component matching pair includes two IMF components. If there is a one-to-many matching situation in DTW, then the closest DTW distance is used as the finally matched IMF component to achieve one-to-one matching and determine several IMF component matching pairs.

[0074] S302. Determine the action difference situation between actions of the same type according to several IMF component matching pairs and the acceleration data.

[0075] Optionally, the action difference situation between actions of the same type includes the first difference situation and the second difference situation. S302 includes steps 1 and 2:

[0076] Step 1. According to several IMF component matching pairs, determine the first difference situation between the actions of each IMF component matching pair under each IMF component.

[0077] First, determine a first action (e.g., the th action) and a second action (e.g., the th action) from various actions respectively. That is, the first action and the second action belong to the same type of action. For each type of action, extract the first action and the second action, and determine a number of candidate IMF component matching pairs corresponding to the first action and the second action according to a number of IMF component matching pairs, that is, all IMF component matching pairs in which the IMF component of the th action is matched with the IMF component of the th action.

[0078] Secondly, determine a first absolute value of the difference in slopes between IMF components in each candidate IMF component matching pair (that is, for the two IMF components in the th pair of candidate IMF component matching pairs, since the IMF component is a part of the PPG heart rate signal, it can also be called the absolute value of the difference in slopes between the signals of the two IMF components in the th pair of candidate IMF component matching pairs).

[0079] Then, respectively determine a first amplitude of the IMF component matched with the IMF component corresponding to the first action in each candidate IMF component matching pair (that is, in the th candidate IMF component matching pair, the amplitude of the IMF component (the signal of the IMF component) matched with the th IMF component corresponding to the th action), a second amplitude of the IMF component matched with the IMF component corresponding to the second action (that is, in the th candidate IMF component matching pair, the amplitude of the IMF component (the signal of the IMF component) matched with the th IMF component corresponding to the th action), and respectively determine a second absolute value of the difference between the first amplitude and the corresponding second amplitude .

[0080] Finally, according to the first absolute value , the second absolute value and the number of candidate IMF component matching pairs, determine, for each type of action and under each IMF component, the first difference situation between the first action and the second action :

[0081]

[0082] Among them, is the The first difference situation between the th action and the th action under a single IMF component. Therefore, the steps of determining the first action and the second action from each type of action are returned until the first difference situation between the actions of each IMF component matching pair is determined for each IMF component. , that is , , When taking different values respectively, represents the first difference situation between the actions of each IMF component matching pair.

[0083] Step 2: Determine the acceleration segment corresponding to each action from the acceleration data, match the acceleration segments corresponding to each action through DTW, respectively determine several acceleration matching pairs between actions of the same type, and determine the second difference situation of the acceleration between the actions of each acceleration matching pair according to the several acceleration matching pairs between actions of the same type.

[0084] First, the acceleration data includes acceleration segments corresponding to several actions respectively. The acceleration segments corresponding to each action are matched through DTW, and several acceleration matching pairs between actions of the same type are respectively determined. The principle of DTW matching is similar to that of IMF component matching and will not be elaborated here. Through matching, several acceleration matching pairs between actions of the same type can be respectively determined. Then, the first action (for example, the th action) and the second action (for example, the th action) are respectively determined from each type of action, and several candidate acceleration matching pairs corresponding to the first action and the second action are determined according to the several acceleration matching pairs between actions of the same type, that is, all the acceleration matching pairs in which the acceleration of the th action is matched with the acceleration of the th action.

[0085] Secondly, determine the third absolute value of the difference in slope between the acceleration segments in each candidate acceleration matching pair (that is, the absolute value of the difference in slope between the acceleration segments in the th pair of candidate acceleration matching pairs).

[0086] Then, respectively determine the third amplitude of the acceleration segment matched with the acceleration segment corresponding to the first action in each candidate acceleration matching pair (that is, in the th candidate acceleration matching pair, the amplitude of the acceleration segment matched with the acceleration segment corresponding to the th action), and the fourth amplitude of the acceleration segment matched with the acceleration segment corresponding to the second action (that is, in the Among the candidate acceleration matching pairs, for the acceleration segment corresponding to the th action, determine the amplitude of the acceleration segment it matches, and respectively determine the fourth absolute value of the difference between the third amplitude and the corresponding fourth amplitude .

[0087] Finally, based on the third absolute value , the fourth absolute value and the number of candidate acceleration matching pairs , determine the second difference situation between the first action and the second action in each type of action:

[0088]

[0089] Among them, is the second difference situation between the th action and the th action. Therefore, return the steps to determine the first action and the second action respectively from each type of action until the second difference situation of the accelerations between the actions of each acceleration matching pair is determined , that is , When taking different values, represents the second difference situation of the accelerations between the actions of each acceleration matching pair.

[0090] S303. According to the temperature data and the action difference situation, determine the variation deviation of each IMF component of each action, and based on the variation deviation of each IMF component of each action, determine the influence situation of each IMF component corresponding to each action on the motion artifacts.

[0091] First, according to the temperature data, determine the temperature DTW matching distance between actions of the same type (that is, the temperature DTW matching distance between the th action and the th action in the same type), and respectively according to the natural exponential function and the temperature DTW matching distance, determine the temperature difference situation between actions of the same type:

[0092]

[0093] Among them, is the temperature difference situation between the th action and the th action in the same type.

[0094] Secondly, respectively determine the ratio of the first difference situation to the second difference situation , and determine the corresponding variance according to each ratio 。

[0095] Then, according to the ratio , the temperature difference and the number of actions in various actions perform a summation process to obtain a summation result . According to the product of the reciprocal of the variance and the summation result, obtain the variation deviation of each IMF component of each action:

[0096]

[0097] wherein, is the variation deviation of the th IMF component of the th action, represents the variation stability of the acceleration data of the current type. The larger its value, the more chaotic the relationship between the acceleration data and the heart rate change is after the th action. For the current chaotic situation, each action may correspond to different heart rate situations.

[0098] Finally, according to the variation deviation of each IMF component of each action, calculate the difference in the variation deviation between two adjacent actions, and determine the influence of each IMF component corresponding to each action on the motion artifact :

[0099] =

[0100] wherein, represents the difference in the variation deviation between the th action and the adjacent th action, denoted as the influence of the th IMF component corresponding to the th action on the motion artifact. In the embodiments of the present application, considering that the monitoring object is a continuous and continuous change during exercise, and at the same time the corresponding change in its heart rate is similar, so when analyzing and reconstructing the weight subsequently, it is necessary to consider the offset of its components caused by the continuous action change. Therefore, in the embodiments of the present application, calculate the influence of each IMF component corresponding to each action on the motion artifact .

[0101] It should be noted that, in actual situations, there may be some motion artifacts in the heart rate information contained in some IMF components. Therefore, it is necessary to analyze the fluctuations in each obtained IMF component. For the IMF components with relatively large influence on the current fluctuations, their weights can be reduced during reconstruction.

[0102] In one implementation, step S400 includes steps S401 - S403:

[0103] S401. Determine the component weight corresponding to each IMF component according to the influence of each IMF component corresponding to each action on motion artifacts.

[0104] Specifically, according to the influence of each IMF component corresponding to each action on motion artifacts to determine the component weight corresponding to each IMF component representing the overall weight corresponding to the th IMF component.

[0105] S402. Respectively determine the amplitude differences between the corresponding IMF components of adjacent pairs of actions, and respectively determine the reconstruction weight of each IMF component of each action according to the amplitude differences and the component weight corresponding to each IMF component.

[0106] Optionally, first, respectively determine the amplitude differences between the corresponding IMF components of adjacent pairs of actions , being the amplitude corresponding to the th IMF component (or the signal of the IMF component) of the th action, being the amplitude corresponding to the th IMF component (or the signal of the IMF component) of the

[0107] a - 1th action. Secondly, respectively determine the fluctuation weight of each IMF component according to the natural exponential function and the opposite of the amplitude difference.

[0108] Then, respectively determine the reconstruction weight of each IMF component of each action according to the product of the fluctuation weight and the component weight of each IMF component:

[0109]

[0110] where is the reconstruction weight of the th IMF component of the th action, being a hyperparameter to prevent the denominator from being zero.

[0111] S403. Perform signal reconstruction according to the reconstruction weights of each IMF component of each action to obtain the target PPG heart rate signal with motion artifacts removed.

[0112] Specifically, according to the reconstruction weights of the respective IMF components of each action, the respective IMF components corresponding to each action are weighted and summed to obtain the target PPG heart rate signal with motion artifacts removed. It can be understood that after the PPG heart rate signal is obtained in real time each time, the target PPG heart rate signal with motion artifacts removed can be finally obtained based on the above process, avoiding the distortion of the heart rate signal caused by motion artifacts, which is beneficial to improving the accuracy of heart rate measurement.

[0113] In one implementation, according to the above operations, the target PPG heart rate signal with motion artifacts removed can be obtained, that is, the target PPG heart rate signal with at least part or all of the motion artifacts removed. At this time, the peak region in the target PPG heart rate signal can be extracted, and then the time interval between two peaks in the peak region can be calculated. Dividing 60 by this time interval can obtain the heart rate. Among them, when displaying the heart rate, the continuous five heart rates can be averaged, and the averaged heart rate is the heart rate value at the current moment.

[0114] The method of the embodiment of the present application decomposes and classifies the actions of the monitored object, analyzes the PPG heart rate signals corresponding to the same type of actions, and uses the influence of the IMF components corresponding to the different classified actions obtained by EMD decomposition on the motion artifacts as the degree of artifact interference in the current PPG heart rate signal. Then, the reconstruction weights of the decomposed IMF components are determined for signal reconstruction to obtain the target PPG heart rate signal with motion artifacts removed, reducing the influence of motion artifacts on the heart rate signal and having an obvious artifact removal effect when the temperature changes.

[0115] It should be noted that the above sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some implementations, multitasking and parallel processing are also possible or may be beneficial.

[0116] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for eliminating motion artifacts from heart rate signals for thermal endurance monitoring, characterized in that, The method includes: During the heat endurance monitoring, acquiring the temperature data, acceleration data, and PPG heart rate signal of blood vessels of the monitored object; Determining each action of different types according to the acceleration data, decomposing the segments of the PPG heart rate signal corresponding to each action of different types, and determining each IMF component corresponding to each action; Determining the influence of each IMF component corresponding to each action on motion artifacts according to each IMF component corresponding to each action and the temperature data; Performing signal reconstruction according to the influence of each IMF component corresponding to each action on motion artifacts to obtain a target PPG heart rate signal with motion artifacts removed; Among them, the method for obtaining the target PPG heart rate signal is: determining the component weight corresponding to each IMF component according to the influence of each IMF component corresponding to each action on motion artifacts; respectively determining the amplitude difference between each IMF component of adjacent two actions, and respectively determining the reconstruction weight of each IMF component of each action according to the amplitude difference and the component weight corresponding to each IMF component; performing signal reconstruction according to the reconstruction weight of each IMF component of each action to obtain a target PPG heart rate signal with motion artifacts removed; Among them, the method for obtaining the reconstruction weight is: respectively determining the fluctuation weight of each IMF component according to the natural exponential function and the opposite number of the amplitude difference; respectively determining the reconstruction weight of each IMF component of each action according to the product of the fluctuation weight of each IMF component and the component weight; Among them, performing signal reconstruction according to the reconstruction weight of each IMF component of each action to obtain a target PPG heart rate signal with motion artifacts removed includes: performing weighted summation on each IMF component corresponding to each action according to the reconstruction weight of each IMF component of each action to obtain a target PPG heart rate signal with motion artifacts removed.

2. The method for eliminating motion artifacts of heart rate signals for thermal endurance monitoring according to claim 1, wherein: The determining each action of different types according to the acceleration data includes: Performing motion decomposition on the acceleration data to obtain motion segments of several actions, and respectively performing normalization processing on each motion segment; Determining the DTW distance between each normalized motion segment, and clustering each motion segment according to the DTW distance to determine each action of different types.

3. The method for eliminating motion artifacts of heart rate signals for thermal endurance monitoring according to claim 1, wherein: The determining the influence of each IMF component corresponding to each action on motion artifacts according to each IMF component corresponding to each action and the temperature data includes: Matching each IMF component corresponding to each action through DTW to respectively determine several IMF component matching pairs between actions of the same type; Determining the action difference situation between actions of the same type according to several IMF component matching pairs and the acceleration data; Based on the temperature data and the action difference situation, determine the change deviation of each IMF component of each action, and based on the change deviation of each IMF component of each action, determine the influence of each IMF component corresponding to each action on motion artifacts.

4. The method for eliminating motion artifacts of a heart rate signal for heat endurance monitoring according to claim 3, characterized in that: The determination of the action difference situation between actions of the same type according to a plurality of the IMF component matching pairs and the acceleration data includes: According to a plurality of the IMF component matching pairs, determine, under each IMF component, the first difference situation between the actions of each of the IMF component matching pairs; Determine the acceleration segment corresponding to each action from the acceleration data, match the acceleration segments corresponding to each action through DTW, respectively determine a plurality of acceleration matching pairs between actions of the same type, and according to the plurality of acceleration matching pairs between actions of the same type, determine the second difference situation of the acceleration between the actions of each of the acceleration matching pairs; Among them, the action difference situation between actions of the same type includes the first difference situation and the second difference situation.

5. The method for eliminating motion artifacts of a heart rate signal for thermal endurance monitoring according to claim 4, wherein: The determination of the first difference situation between the actions of each of the IMF component matching pairs under each IMF component according to a plurality of the IMF component matching pairs includes: Respectively determine a first action and a second action from each type of action, and according to a plurality of the IMF component matching pairs, determine a plurality of candidate IMF component matching pairs corresponding to the first action and the second action; Determine the first absolute value of the difference in slope between the IMF components in each of the candidate IMF component matching pairs; Respectively determine the first amplitude of the IMF component matched with the IMF component corresponding to the first action and the second amplitude of the IMF component matched with the IMF component corresponding to the second action in each of the candidate IMF component matching pairs, and respectively determine the second absolute value of the difference between the first amplitude and the corresponding second amplitude; According to the first absolute value, the second absolute value, and the number of the candidate IMF component matching pairs, determine, in each type of action, the first difference situation between the first action and the second action under each IMF component, and return to the step of respectively determining the first action and the second action from each type of action until the first difference situation between the actions of each of the IMF component matching pairs under each IMF component is determined.

6. The method for eliminating motion artifacts of a heart rate signal for thermal endurance monitoring according to claim 4, characterized in that: The determination of the second difference situation of the acceleration between the actions of each of the acceleration matching pairs according to a plurality of the acceleration matching pairs between actions of the same type includes: Respectively determine a first action and a second action from each type of action, and according to a plurality of the acceleration matching pairs between actions of the same type, determine a plurality of candidate acceleration matching pairs corresponding to the first action and the second action; Determine the third absolute value of the difference in slope between the acceleration segments in each of the candidate acceleration matching pairs; For each of the candidate acceleration matching pairs, determine the third amplitude of the acceleration segment that matches the acceleration segment corresponding to the first action, and the fourth amplitude of the acceleration segment that matches the acceleration segment corresponding to the second action, and respectively determine the fourth absolute value of the difference between the third amplitude and the corresponding fourth amplitude; Based on the fourth absolute value, the fourth absolute value, and the number of candidate acceleration matching pairs, determine the second difference situation between the first action and the second action in each type of action, and return the step of respectively determining the first action and the second action from each type of action until the second difference situation of the acceleration between the actions of each of the acceleration matching pairs is determined.

7. The method for eliminating motion artifacts of a heart rate signal for thermal endurance monitoring according to claim 4, characterized in that: The determining, according to the temperature data and the action difference situation, the change deviation situation of each IMF component of each action, and according to the change deviation situation of each IMF component of each action, the influence situation of each IMF component corresponding to each action on motion artifacts includes: According to the temperature data, determine the temperature DTW matching distance between actions of the same type, and respectively determine the temperature difference situation between actions of the same type according to the natural exponential function and the temperature DTW matching distance; Respectively determine the ratio of the first difference situation to the second difference situation, and determine the corresponding variance according to each of the ratios; Perform a summation process according to the ratio, the temperature difference situation, and the number of actions in each type of action to obtain a summation result, and obtain the change deviation situation of each IMF component of each action according to the product of the reciprocal of the variance and the summation result; According to the change deviation situation of each IMF component of each action, calculate the difference in the change deviation situation between two adjacent actions, and determine the influence situation of each IMF component corresponding to each action on motion artifacts.

Images