CN118105051B - Rehabilitation cloud platform system for monitoring cardiopulmonary function - Google Patents
Rehabilitation cloud platform system for monitoring cardiopulmonary function Download PDFInfo
- Publication number
- CN118105051B CN118105051B CN202410534352.9A CN202410534352A CN118105051B CN 118105051 B CN118105051 B CN 118105051B CN 202410534352 A CN202410534352 A CN 202410534352A CN 118105051 B CN118105051 B CN 118105051B
- Authority
- CN
- China
- Prior art keywords
- heart rate
- moment
- data
- predicted
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 47
- 230000002612 cardiopulmonary effect Effects 0.000 title claims abstract description 17
- 230000008859 change Effects 0.000 claims abstract description 18
- 238000003062 neural network model Methods 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims abstract description 16
- 238000004458 analytical method Methods 0.000 claims abstract description 12
- 230000000694 effects Effects 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 10
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 39
- 239000008280 blood Substances 0.000 claims description 39
- 210000004369 blood Anatomy 0.000 claims description 39
- 229910052760 oxygen Inorganic materials 0.000 claims description 39
- 239000001301 oxygen Substances 0.000 claims description 39
- 230000029058 respiratory gaseous exchange Effects 0.000 claims description 21
- 238000009499 grossing Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 10
- 230000036387 respiratory rate Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 230000035565 breathing frequency Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 abstract description 20
- 230000004199 lung function Effects 0.000 abstract description 16
- 230000004217 heart function Effects 0.000 abstract description 14
- 230000000875 corresponding effect Effects 0.000 description 22
- 230000008569 process Effects 0.000 description 12
- 230000002159 abnormal effect Effects 0.000 description 11
- 230000005856 abnormality Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 4
- 230000007423 decrease Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 230000000241 respiratory effect Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000036391 respiratory frequency Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Abstract
The invention relates to the technical field of neural network model monitoring, in particular to a rehabilitation cloud platform system for monitoring cardiopulmonary function. The system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring heart rate data and other physical characteristic data affecting heart rate variation and determining the noise degree of the heart rate data at each moment; the heart rate prediction module is used for determining a prediction error coefficient of each moment except the first moment, and the analysis module is used for constructing a time sequence heart rate fluctuation model and determining a predicted heart rate of the moment next to the current moment based on the neural network model and the time sequence heart rate fluctuation model; and the monitoring module is used for determining a corrected predicted heart rate and taking the corrected predicted heart rate as a heart and lung function monitoring result. The heart-lung function monitoring method and device can combine the change of other physical characteristic data, the neighborhood numerical fluctuation of heart rate data and the fluctuation analysis of whole heart rate data, improve the accuracy of correcting and predicting heart rate, improve the objectivity and reliability of heart-lung function monitoring results and enhance the heart-lung function monitoring effect.
Description
Technical Field
The invention relates to the technical field of neural network model monitoring, in particular to a rehabilitation cloud platform system for monitoring cardiopulmonary function.
Background
The monitoring of the physical health condition is a health care mode which is gradually popularized in society, the monitoring of the heart and lung functions is an important part of the physical health condition, and in the process of monitoring the heart and lung functions, the future heart rate is predicted, so that effective danger early warning can be carried out, and the heart and lung functions are the main functions of the heart and lung functions prediction.
In the related art, data analysis is realized through numerical variation of heart rate data, and heart rate prediction is performed at future time, in this way, heart rate variation is influenced by various conditions, such as physique of a tester, exercise amount at current time and the like, so that heart rate prediction accuracy is insufficient, and heart and lung function monitoring effect is poor.
Disclosure of Invention
In order to solve the technical problems of insufficient heart rate prediction accuracy and poor heart lung function monitoring effect caused by data analysis and heart rate prediction only through the numerical value of heart rate data in the related art, the invention provides a rehabilitation cloud platform system for heart lung function monitoring, which adopts the following specific technical scheme:
The invention provides a rehabilitation cloud platform system for monitoring cardiopulmonary function, which comprises:
The acquisition module is used for acquiring heart rate data and other physique feature data affecting heart rate variation, and determining the noise degree of the heart rate data at each moment according to the variation of the other physique feature data and the variation of the heart rate data at different moments, wherein the other physique feature data comprises respiratory rate and blood oxygen saturation;
The heart rate prediction module is used for predicting heart rate data of the next moment according to the heart rate data of any moment and the corresponding previous moment to obtain predicted data of the next moment, and determining a predicted error coefficient of the next moment according to the predicted data of the next moment and the real heart rate data of the next moment, so that the predicted error coefficient of each moment except the first moment is obtained;
The analysis module is used for determining the time to be analyzed with the prediction influence on the current time according to the prediction error coefficients of all the times except the first time; constructing a time sequence heart rate fluctuation model according to heart rate data of all to-be-analyzed moments, and predicting the heart rate data of the next moment at the current moment based on the neural network model and the time sequence heart rate fluctuation model to obtain a predicted heart rate of the next moment at the current moment;
The monitoring module is used for correcting the predicted heart rate at the next moment of the current moment according to the heart rate data and the noise degree of all moments of the current moment in the preset time sequence range to obtain a corrected predicted heart rate, and the corrected predicted heart rate is used as a heart-lung function monitoring result.
Further, the determining the noise degree of the heart rate data at each moment according to the change of the other physical characteristic data and the change of the heart rate data at different moments comprises:
taking any moment as a moment to be measured, and taking the other two moments which are closest to each other in time sequence as neighborhood moments;
Calculating the difference value of heart rate data between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as heart rate variation index to be measured at the moment to be measured; calculating the difference value of the breathing frequency between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as the breathing variation index to be measured at the moment to be measured; calculating the difference value of the blood oxygen saturation between the time to be measured and each neighborhood time, and obtaining and normalizing the difference value to be measured as the blood oxygen variation index to be measured at the time to be measured;
calculating the absolute value of the difference between the heart rate variation index to be measured and the respiration variation index to be measured to obtain the heart rate respiration variation difference; calculating the absolute value of the difference between the heart rate variation index to be measured and the blood oxygen variation index to be measured to obtain the heart rate blood oxygen variation difference;
And determining the noise degree of the heart rate data at the moment to be detected according to the heart rate respiration variation difference and the heart rate blood oxygen variation difference, and updating the moment to be detected to obtain the noise degree of the heart rate data at each moment.
Further, the difference of the heart rate and the respiration changes and the noise degree are in positive correlation, the difference of the heart rate and the blood oxygen changes and the noise degree are in negative correlation, and the value of the noise degree is a normalized value.
Further, predicting the heart rate data at the next moment according to the heart rate data at any moment and the corresponding previous moment to obtain predicted data at the next moment, including:
And carrying out exponential smoothing calculation on heart rate data at any moment and corresponding previous moment according to an exponential smoothing algorithm based on a preset weight value to obtain predicted data at the next moment.
Further, the determining the prediction error coefficient of the next moment according to the prediction data of the next moment and the real heart rate data of the next moment includes:
and calculating the absolute value of the difference between the predicted data at the next moment and the real heart rate data at the next moment, and obtaining a predicted error coefficient at the next moment through normalization processing.
Further, the determining, according to the prediction error coefficients of all the moments except the first moment, the moment to be analyzed having a prediction effect on the current moment includes:
And taking the time when the prediction error coefficient is smaller than a preset error threshold value as the time to be analyzed.
Further, the constructing a time sequence heart rate fluctuation model according to the heart rate data of all the to-be-analyzed moments includes:
And constructing a two-dimensional coordinate system by taking the time sequence as an abscissa and taking heart rate data as an ordinate, determining coordinate points of heart rate data of all moments to be analyzed in the two-dimensional coordinate system, and connecting two coordinate points adjacent to the time sequence to obtain a time sequence heart rate fluctuation model.
Further, the predicting the heart rate data at the next moment in the current moment based on the neural network model and the time sequence heart rate fluctuation model to obtain the predicted heart rate at the next moment in the current moment comprises the following steps:
Inputting all coordinate points in the time sequence heart rate fluctuation model into a pre-trained neural network model, processing through the neural network model, and outputting a predicted heart rate at the next moment of the current moment.
Further, the correcting the predicted heart rate at the next time of the current time according to the heart rate data and the noise degree of all the times in the preset time sequence range at the current time to obtain a corrected predicted heart rate includes:
performing negative correlation mapping on all the noise degrees to obtain the influence weights of the heart rate data at each moment, wherein the sum value of the influence weights of the heart rate data at all the moments in a preset time sequence range is 1;
Calculating the product of heart rate data and influence weights at each moment to obtain heart rate influence values at each moment, and taking the sum of heart rate influence values at all moments in a preset time sequence range as a neighborhood heart rate prediction value;
And determining the corrected predicted heart rate at the next moment of the current moment according to the predicted heart rate at the next moment of the current moment and the neighborhood heart rate predicted value.
Further, the determining the corrected predicted heart rate at the next moment of the current moment according to the predicted heart rate at the next moment of the current moment and the neighborhood heart rate predicted value comprises:
and calculating the average value of the predicted heart rate and the neighborhood heart rate predicted value at the next moment of the current moment as the corrected predicted heart rate.
The invention has the following beneficial effects:
Aiming at the problems of insufficient accuracy of heart rate prediction and poor monitoring effect of heart and lung functions caused by abnormal analysis and heart rate prediction only through the numerical value of heart rate data in the related technology, the noise degree of the heart rate data at each moment is analyzed through the heart rate data and other physique characteristic data such as the respiratory frequency, the blood oxygen saturation and the like which influence heart rate variation, and the objective variation of the heart rate data can be analyzed by combining with the related multi-type other physique characteristic data, so that the objectivity of heart rate data analysis is improved, and the real heart rate variation characteristics are more met; by analyzing the prediction error coefficients at different moments, the abnormality of each moment can be ensured to be analyzed, the moment to be analyzed with prediction influence on the current moment is further screened out, the moment with larger abnormality can be screened out by combining the characteristic that the noise causes data change suddenly, the accuracy of the subsequent heart rate prediction is further ensured, and the predicted heart rate at the moment next to the current moment is obtained; and then, correcting the predicted heart rate by combining heart rate data and noise degrees of all the moments of the current moment in a preset time sequence range of the local neighborhood to obtain corrected predicted heart rate, so that the accuracy of correcting the predicted heart rate can be improved by combining the change of other physical characteristic data, neighborhood numerical fluctuation of heart rate data and fluctuation analysis of whole heart rate data, and the corrected predicted heart rate is used as a heart and lung function monitoring result, thereby enabling the heart and lung function monitoring result to have stronger objectivity and reliability, reducing the influence of abnormal data and noise and enhancing the heart and lung function monitoring effect.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a diagram of a rehabilitation cloud platform system for monitoring cardiopulmonary function according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of a rehabilitation cloud platform system for monitoring cardiopulmonary function according to the invention, which is specific implementation, structure, characteristics and effects thereof, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of a rehabilitation cloud platform system for monitoring cardiopulmonary function provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a structural diagram of a rehabilitation cloud platform system for monitoring cardiopulmonary function according to an embodiment of the present invention is shown, where the system includes:
The obtaining module 101 is configured to obtain heart rate data and other physical characteristic data affecting heart rate variation, and determine a noise level of the heart rate data at each moment according to the variation of the other physical characteristic data and the variation of the heart rate data at different moments, where the other physical characteristic data includes a respiratory rate and a blood oxygen saturation level.
The invention has the specific implementation scene that based on the heart rate data and the changes of the respiratory rate, the blood oxygen saturation and the like, the future heart rate data is predicted based on the changes of the heart rate data at other moments in a time sequence neighborhood range, so that the invention is convenient for realizing rehabilitation monitoring and early warning by combining the characteristic performance of relevant staff to actual testing staff according to the prediction result.
In the embodiment of the invention, a portable heart rate detector or intelligent equipment with a portable heart rate detection function, such as a smart watch, a smart phone and the like, can be used for acquiring heart rate data of a tester at different moments, and of course, the respiratory rate and the blood oxygen saturation can be acquired by adopting corresponding equipment. It should be noted that the collection period of the heart rate data, the respiration rate and the blood oxygen saturation level is the same, for example, the heart rate data, the respiration rate and the blood oxygen saturation level may be collected every one minute, or a plurality of other times may be preset, for example, the heart rate data, the respiration rate and the blood oxygen saturation level may be collected every 10 seconds. That is, the heart rate data, the respiratory rate and the blood oxygen saturation are required to be acquired at the same time to avoid errors caused by mismatching of acquisition of different types of data in time sequence.
It should be noted that, because the heart rate data inevitably generates a certain noise influence in the process of collection, storage and transmission, the obtained heart rate data is directly used for prediction without considering the noise influence, which may result in lower prediction accuracy, and the changes of the heart rate data, the respiratory rate, the blood oxygen saturation and the like can analyze the noise.
Further, in some embodiments of the present invention, determining the noise level of the heart rate data at each time according to the change of the other physical characteristic data and the change of the heart rate data at different times includes: taking any moment as a moment to be measured, and taking the other two moments which are closest to each other in time sequence as neighborhood moments; calculating the difference value of heart rate data between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as heart rate variation index to be measured at the moment to be measured; calculating the difference value of the breathing frequency between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as the breathing variation index to be measured at the moment to be measured; calculating the difference value of the blood oxygen saturation between the time to be measured and each neighborhood time, and obtaining and normalizing the difference value to be measured as the blood oxygen variation index to be measured at the time to be measured; calculating the absolute value of the difference between the heart rate variation index to be measured and the respiration variation index to be measured to obtain the heart rate respiration variation difference; calculating the absolute value of the difference between the heart rate variation index to be measured and the blood oxygen variation index to be measured to obtain the heart rate blood oxygen variation difference; and determining the noise degree of the heart rate data at the moment to be detected according to the heart rate respiration variation difference and the heart rate blood oxygen variation difference, and updating the moment to be detected to obtain the noise degree of the heart rate data at each moment.
When the heart rate is high, the corresponding respiratory rate is high, and the blood oxygen saturation is correspondingly low, so that in the embodiment of the invention, the time sequence change of the data is analyzed by combining the objective characteristics.
In the embodiment of the invention, the neighborhood time is two times closest to the time to be measured, and when the time to be measured is not the first time and the last time, the corresponding times are the times on the left side and the right side of the time to be measured, that is, the time to be measured can be subjected to change analysis according to the data of the neighborhood time.
The difference between the heart rate data at the moment to be measured and each neighborhood moment is calculated, the difference is averaged and normalized to serve as a heart rate variation index to be measured at the moment to be measured, namely, the average value of the difference between the two heart rate data is calculated, then the influence of dimension is eliminated by normalization to obtain the heart rate variation index to be measured, the respiration variation index to be measured and the blood oxygen variation index to be measured are similarly calculated, the calculation is carried out according to the variation of the corresponding data at the neighborhood moment and the moment to be measured, and the objective variation of heart rate, respiration frequency and blood oxygen saturation is combined to realize noise analysis.
In the embodiment of the invention, the absolute values of the difference values of the heart rate variation index to be measured, the blood oxygen variation index to be measured and the respiratory variation index to be measured are respectively calculated to obtain the corresponding heart rate blood oxygen variation difference and heart rate respiratory variation difference, and as the heart rate is faster, the corresponding respiratory rate is faster, the heart rate is positively correlated with the respiratory rate variation, the variation under normal condition is smaller, the blood oxygen saturation is correspondingly reduced, the heart rate is negatively correlated with the blood oxygen saturation variation, and the variation under normal condition is larger, so that the smaller the value of the heart rate respiratory variation difference is, the larger the value of the heart rate blood oxygen variation difference is, and the corresponding noise influence is smaller.
Therefore, in the embodiment of the invention, the heart rate respiration variation difference and the noise degree are in positive correlation, the heart rate blood oxygen variation difference and the noise degree are in negative correlation, and the value of the noise degree is a normalized numerical value.
In one embodiment of the present invention, the normalization process may specifically be, for example, maximum and minimum normalization processes, and the normalization in the subsequent steps may be performed by using the maximum and minimum normalization processes, and in other embodiments of the present invention, other normalization methods may be selected according to a specific range of values, which will not be described herein.
The positive correlation relationship indicates that the dependent variable increases along with the increase of the independent variable, the dependent variable decreases along with the decrease of the independent variable, and the specific relationship can be multiplication relationship, addition relationship, idempotent of an exponential function and is determined by practical application; the negative correlation indicates that the dependent variable decreases with increasing independent variable, and the dependent variable increases with decreasing independent variable, which may be a subtraction relationship, a division relationship, or the like, and is determined by the actual application.
The invention can calculate the difference between the difference of the heart rate respiration variation and the difference of the heart rate blood oxygen variation, and the noise degree is obtained by normalization processing.
The heart rate prediction module 102 is configured to predict heart rate data at a next time according to any time and heart rate data corresponding to a previous time, obtain predicted data at the next time, and determine a prediction error coefficient at the next time according to the predicted data at the next time and real heart rate data at the next time, thereby obtaining a prediction error coefficient at each time except the first time.
Further, in some embodiments of the present invention, predicting heart rate data at a next time according to heart rate data at any time and corresponding previous time to obtain predicted data at the next time includes: and carrying out exponential smoothing calculation on heart rate data at any moment and corresponding previous moment according to an exponential smoothing algorithm based on a preset weight value to obtain predicted data at the next moment.
The preset weight is a weight corresponding to exponential smoothing, optionally, the preset weight may be specifically, for example, 0.5, or may be adjusted according to an actual detection requirement, which is not limited.
In the embodiment of the present invention, the predicted data at the next time is specifically the heart rate data predicted at the next time, and it should be noted that, because the noise influence at each time is not considered, the accuracy of the predicted data is low, and then the predicted data needs to be specifically corrected.
The formula for calculating the predicted data at the next moment by using the exponential smoothing algorithm in the invention can be specifically: In the method, in the process of the invention, Predictive data representing the next time to the c-th time, a representing a preset weight,The predicted data representing the c-th instant,Actual heart rate data at time c is indicated.
It should be noted that, the exponential smoothing algorithm is a weighted moving smoothing algorithm well known in the art, which is not further limited and described herein. The prediction data of the next moment is obtained by weighting the prediction data of the previous moment and the actual data in the formula and predicting the next moment.
Further, in some embodiments of the present invention, determining the prediction error coefficient at the next time based on the prediction data at the next time and the actual heart rate data at the next time includes: and calculating the absolute value of the difference between the predicted data at the next moment and the real heart rate data at the next moment, and obtaining a predicted error coefficient at the next moment through normalization processing.
In the embodiment of the invention, the prediction error coefficient of the next moment is obtained through the difference between the prediction data and the real data of the next moment, can be used as the difference between heart rate variation in the prediction and the actual process, and can be used for effectively comparing the prediction value and the real value in the prediction process of the next moment to obtain the prediction error coefficient, wherein the prediction error coefficient characterizes the abnormal condition in the prediction process.
It should be noted that, the abnormal situation may be specifically caused by, for example, the influence of noise in the data transmission process on the real data at the previous moment, or the heart rate variation in the abnormal scene is abnormal, so the present invention screens the abnormal situation in the subsequent embodiment, specifically referring to the subsequent embodiment.
An analysis module 103, configured to determine a time to be analyzed having a predicted impact on a current time according to prediction error coefficients of all the times except the first time; and constructing a time sequence heart rate fluctuation model according to heart rate data of all the moments to be analyzed, and predicting the heart rate data of the moment next to the current moment based on the neural network model and the time sequence heart rate fluctuation model to obtain a predicted heart rate of the moment next to the current moment.
Further, in some embodiments of the present invention, determining a time to be analyzed having a predicted impact on a current time according to prediction error coefficients of all times except the first time includes: and taking the time when the prediction error coefficient is smaller than the preset error threshold value as the time to be analyzed.
In the embodiment of the invention, the larger the prediction error coefficient is, the larger the difference between the heart rate fluctuation at the corresponding moment to be analyzed and the heart rate fluctuation at other moments in the local neighborhood range is, namely, the higher the abnormal degree of heart rate data at the moment to be analyzed is, so that the threshold value is set for screening data, and the normal moment to be analyzed of the heart rate data is obtained through screening. Alternatively, in some embodiments of the present invention, the preset error threshold may be specifically, for example, 0.6, and of course, in other embodiments of the present invention, the preset error threshold may be adjusted according to the actual detection requirement, which is not limited.
Further, in some embodiments of the present invention, constructing a time-series heart rate fluctuation model according to heart rate data of all the time points to be analyzed includes: and constructing a two-dimensional coordinate system by taking the time sequence as an abscissa and taking heart rate data as an ordinate, determining coordinate points of heart rate data of all moments to be analyzed in the two-dimensional coordinate system, and connecting two coordinate points adjacent to the time sequence to obtain a time sequence heart rate fluctuation model.
In the embodiment of the invention, because the time to be analyzed is the time with higher overall normal performance of heart rate data, the data prediction is performed based on all the time to be analyzed, and the influence of abnormal time can be effectively reduced.
Further, in some embodiments of the present invention, predicting heart rate data at a next moment in a current moment based on a neural network model and a time-series heart rate fluctuation model to obtain a predicted heart rate at the next moment in the current moment includes: all coordinate points in the time sequence heart rate fluctuation model are input into a pre-trained neural network model, and the predicted heart rate at the next moment of the current moment is output through the neural network model processing.
The neural network model in the embodiment of the invention can specifically use a single hidden layer network, use relu as an activation function, further optimize the neural network based on modes such as moving average, regularization and exponential decay, and the like, and it is noted that the structure of the neural network can generate corresponding changes when the neural network is constructed in the face of different data sets, different scenes and different prediction demands, and the structure of the neural network and the subsequent pre-training process are all technologies well known to those skilled in the art, and are not further limited and repeated.
In the embodiment of the invention, the data prediction can be performed based on the time sequence heart rate fluctuation model, and the neural network model is pre-trained, so that all coordinate points in the time sequence heart rate fluctuation model are input, and the predicted heart rate is output.
The monitoring module 104 is configured to correct the predicted heart rate at the next moment of the current moment according to the heart rate data and the noise level at all moments of the current moment within the preset time sequence range, to obtain a corrected predicted heart rate, and take the corrected predicted heart rate as a monitoring result of the cardiopulmonary function.
Further, in some embodiments of the present invention, correcting the predicted heart rate at the next time of the current time according to the heart rate data and the noise level at all times of the current time within the preset time sequence range to obtain a corrected predicted heart rate includes: performing negative correlation mapping on all noise degrees to obtain influence weights of heart rate data at each moment, wherein the sum of the influence weights of the heart rate data at all moments in a preset time sequence range is 1; calculating the product of heart rate data and influence weights at each moment to obtain heart rate influence values at each moment, and taking the sum of heart rate influence values at all moments in a preset time sequence range as a neighborhood heart rate prediction value; and determining the corrected predicted heart rate at the next moment of the current moment according to the predicted heart rate at the next moment of the current moment and the neighborhood heart rate predicted value.
The preset time sequence range is, for example, a time sequence range corresponding to the first 10 times of the current time, which is optionally not limited, for example.
In the embodiment of the invention, as the noise degree characterizes the noise influence at the corresponding moment, namely the change of heart rate data does not accord with the normal change, the noise degree is subjected to negative correlation mapping, and the obtained influence weight has positive correlation on the heart rate normality. That is, the less the heart rate data is affected by noise, the less the corresponding impact weight.
Obtaining a heart rate influence value of a corresponding moment through the product of heart rate data and influence weights of each moment, then taking the sum value of the heart rate influence values of all the moments in a preset time sequence range as a neighborhood heart rate prediction value, wherein the prediction value combines the heart rates of all the moments in the neighborhood range and the corresponding influence weights, and the obtained prediction value is characterized in that the heart rate change in time sequence is smaller, and the heart rate change of each minute is in linear change or in gentle fluctuation, so that the heart rate data of the next moment is subjected to the current and the neighborhood influence is larger, and the heart rate prediction method analyzes the heart rate to obtain a corrected predicted heart rate.
Further, in some embodiments of the present invention, determining a corrected predicted heart rate at a time next to the current time based on the predicted heart rate at the time next to the current time and the neighborhood heart rate prediction value includes: and calculating the average value of the predicted heart rate and the neighborhood heart rate predicted value at the next moment of the current moment as the corrected predicted heart rate.
Of course, in other embodiments of the present invention, corresponding weight values may be respectively assigned to the predicted heart rate and the neighborhood heart rate predicted value at the next moment of the current moment, so that a corrected predicted heart rate is obtained by performing weighted analysis according to the weight values.
In the embodiment of the invention, the corrected and predicted heart rate is used as the heart-lung function monitoring result, so that the related staff can combine the heart-lung function monitoring result and other characteristics of the testers in the follow-up process to realize heart-lung function monitoring.
Aiming at the problems of insufficient accuracy of heart rate prediction and poor monitoring effect of heart and lung functions caused by abnormal analysis and heart rate prediction only through the numerical value of heart rate data in the related technology, the noise degree of the heart rate data at each moment is analyzed through the heart rate data and other physique characteristic data such as the respiratory frequency, the blood oxygen saturation and the like which influence heart rate variation, and the objective variation of the heart rate data can be analyzed by combining with the related multi-type other physique characteristic data, so that the objectivity of heart rate data analysis is improved, and the real heart rate variation characteristics are more met; by analyzing the prediction error coefficients at different moments, the abnormality of each moment can be ensured to be analyzed, the moment to be analyzed with prediction influence on the current moment is further screened out, the moment with larger abnormality can be screened out by combining the characteristic that the noise causes data change suddenly, the accuracy of the subsequent heart rate prediction is further ensured, and the predicted heart rate at the moment next to the current moment is obtained; and then, correcting the predicted heart rate by combining heart rate data and noise degrees of all the moments of the current moment in a preset time sequence range of the local neighborhood to obtain corrected predicted heart rate, so that the accuracy of correcting the predicted heart rate can be improved by combining the change of other physical characteristic data, neighborhood numerical fluctuation of heart rate data and fluctuation analysis of whole heart rate data, and the corrected predicted heart rate is used as a heart and lung function monitoring result, thereby enabling the heart and lung function monitoring result to have stronger objectivity and reliability, reducing the influence of abnormal data and noise and enhancing the heart and lung function monitoring effect.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (7)
1. A rehabilitation cloud platform system for cardiopulmonary function monitoring, the system comprising:
The acquisition module is used for acquiring heart rate data and other physique feature data affecting heart rate variation, and determining the noise degree of the heart rate data at each moment according to the variation of the other physique feature data and the variation of the heart rate data at different moments, wherein the other physique feature data comprises respiratory rate and blood oxygen saturation;
The heart rate prediction module is used for predicting heart rate data of the next moment according to the heart rate data of any moment and the corresponding previous moment to obtain predicted data of the next moment, and determining a predicted error coefficient of the next moment according to the predicted data of the next moment and the real heart rate data of the next moment, so that the predicted error coefficient of each moment except the first moment is obtained;
The analysis module is used for determining the time to be analyzed with the prediction influence on the current time according to the prediction error coefficients of all the times except the first time; constructing a time sequence heart rate fluctuation model according to heart rate data of all to-be-analyzed moments, and predicting the heart rate data of the next moment at the current moment based on the neural network model and the time sequence heart rate fluctuation model to obtain a predicted heart rate of the next moment at the current moment;
The monitoring module is used for correcting the predicted heart rate at the next moment of the current moment according to the heart rate data and the noise degree of all moments of the current moment in a preset time sequence range to obtain a corrected predicted heart rate, and the corrected predicted heart rate is used as a heart-lung function monitoring result;
the determining the noise degree of the heart rate data at each moment according to the change of the other physical characteristic data and the change of the heart rate data at different moments comprises the following steps:
taking any moment as a moment to be measured, and taking the other two moments which are closest to each other in time sequence as neighborhood moments;
Calculating the difference value of heart rate data between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as heart rate variation index to be measured at the moment to be measured; calculating the difference value of the breathing frequency between the moment to be measured and each neighborhood moment, and obtaining and normalizing the difference value to be measured as the breathing variation index to be measured at the moment to be measured; calculating the difference value of the blood oxygen saturation between the time to be measured and each neighborhood time, and obtaining and normalizing the difference value to be measured as the blood oxygen variation index to be measured at the time to be measured;
calculating the absolute value of the difference between the heart rate variation index to be measured and the respiration variation index to be measured to obtain the heart rate respiration variation difference; calculating the absolute value of the difference between the heart rate variation index to be measured and the blood oxygen variation index to be measured to obtain the heart rate blood oxygen variation difference;
Determining the noise degree of heart rate data at the moment to be detected according to the heart rate respiration variation difference and the heart rate blood oxygen variation difference, updating the moment to be detected, and obtaining the noise degree of the heart rate data at each moment, wherein the heart rate respiration variation difference and the noise degree are in positive correlation, the heart rate blood oxygen variation difference and the noise degree are in negative correlation, and the value of the noise degree is a normalized numerical value;
Correcting the predicted heart rate at the next moment of the current moment according to the heart rate data and the noise degree of all moments of the current moment in a preset time sequence range to obtain a corrected predicted heart rate, wherein the method comprises the following steps:
performing negative correlation mapping on all the noise degrees to obtain the influence weights of the heart rate data at each moment, wherein the sum value of the influence weights of the heart rate data at all the moments in a preset time sequence range is 1;
Calculating the product of heart rate data and influence weights at each moment to obtain heart rate influence values at each moment, and taking the sum of heart rate influence values at all moments in a preset time sequence range as a neighborhood heart rate prediction value;
And determining the corrected predicted heart rate at the next moment of the current moment according to the predicted heart rate at the next moment of the current moment and the neighborhood heart rate predicted value.
2. The rehabilitation cloud platform system for monitoring cardiopulmonary function according to claim 1, wherein predicting the heart rate data at the next moment according to the heart rate data at any moment and the corresponding previous moment to obtain the predicted data at the next moment comprises:
And carrying out exponential smoothing calculation on heart rate data at any moment and corresponding previous moment according to an exponential smoothing algorithm based on a preset weight value to obtain predicted data at the next moment.
3. A rehabilitation cloud platform system for cardiopulmonary function monitoring as claimed in claim 2, wherein said determining a prediction error coefficient at the next time based on the prediction data at the next time and the actual heart rate data at the next time comprises:
and calculating the absolute value of the difference between the predicted data at the next moment and the real heart rate data at the next moment, and obtaining a predicted error coefficient at the next moment through normalization processing.
4. A rehabilitation cloud platform system for cardiopulmonary function monitoring according to claim 1, wherein the determining the time to be analyzed having a predicted effect on the current time according to the prediction error coefficients of all the time except the first time comprises:
And taking the time when the prediction error coefficient is smaller than a preset error threshold value as the time to be analyzed.
5. The rehabilitation cloud platform system for monitoring cardiopulmonary function according to claim 1, wherein the constructing a time-series heart rate fluctuation model according to heart rate data of all to-be-analyzed moments comprises:
And constructing a two-dimensional coordinate system by taking the time sequence as an abscissa and taking heart rate data as an ordinate, determining coordinate points of heart rate data of all moments to be analyzed in the two-dimensional coordinate system, and connecting two coordinate points adjacent to the time sequence to obtain a time sequence heart rate fluctuation model.
6. The rehabilitation cloud platform system for monitoring cardiopulmonary function according to claim 5, wherein predicting the heart rate data at the next moment in time based on the neural network model and the time sequence heart rate fluctuation model to obtain the predicted heart rate at the next moment in time comprises:
Inputting all coordinate points in the time sequence heart rate fluctuation model into a pre-trained neural network model, processing through the neural network model, and outputting a predicted heart rate at the next moment of the current moment.
7. The rehabilitation cloud platform system for monitoring cardiopulmonary function according to claim 1, wherein the determining the corrected predicted heart rate at the next moment of the current moment according to the predicted heart rate at the next moment of the current moment and the neighborhood heart rate predicted value comprises:
and calculating the average value of the predicted heart rate and the neighborhood heart rate predicted value at the next moment of the current moment as the corrected predicted heart rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410534352.9A CN118105051B (en) | 2024-04-30 | Rehabilitation cloud platform system for monitoring cardiopulmonary function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410534352.9A CN118105051B (en) | 2024-04-30 | Rehabilitation cloud platform system for monitoring cardiopulmonary function |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118105051A CN118105051A (en) | 2024-05-31 |
| CN118105051B true CN118105051B (en) | 2024-07-02 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013036718A1 (en) * | 2011-09-08 | 2013-03-14 | Isis Innovation Ltd. | Determining acceptability of physiological signals |
| CN105266776A (en) * | 2015-09-06 | 2016-01-27 | 陈才维 | Method for monitoring human health |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013036718A1 (en) * | 2011-09-08 | 2013-03-14 | Isis Innovation Ltd. | Determining acceptability of physiological signals |
| CN105266776A (en) * | 2015-09-06 | 2016-01-27 | 陈才维 | Method for monitoring human health |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110852515B (en) | Water quality index prediction method based on mixed long-time and short-time memory neural network | |
| CN110772700B (en) | Automatic sleep-aiding music pushing method and device, computer equipment and storage medium | |
| CN103020472A (en) | Physiological signal quality evaluation method and system based on constrained estimation | |
| CN111611888B (en) | Non-contact blood pressure estimation device | |
| EP3656304B1 (en) | A method and a system for determining the maximum heart rate of a user in a freely performed physical exercise | |
| CN107945817A (en) | Heart and lung sounds signal sorting technique, detection method, device, medium and computer equipment | |
| CN110600053A (en) | Cerebral stroke dysarthria risk prediction method based on ResNet and LSTM network | |
| CN109645989A (en) | A kind of depth of anesthesia estimation method and system | |
| CN108021888A (en) | A kind of fall detection method | |
| CN113288131B (en) | Non-invasive blood glucose detection method, processor and device based on graph convolution network | |
| CN112057066A (en) | Heart rate detection method, wearable device and computer storage medium | |
| CN103310113A (en) | Universal blood glucose prediction method based on frequency band separation and data modeling | |
| CN113701825A (en) | Body-building facility abnormity detection method and system based on artificial intelligence | |
| CN116821809A (en) | Vital sign data acquisition system based on artificial intelligence | |
| Chen et al. | Edge2Analysis: a novel AIoT platform for atrial fibrillation recognition and detection | |
| CN116842460A (en) | Cough-related disease identification method and system based on attention mechanism and residual neural network | |
| KR101603308B1 (en) | Biological age calculation model generation method and system thereof, biological age calculation method and system thereof | |
| CN118105051B (en) | Rehabilitation cloud platform system for monitoring cardiopulmonary function | |
| CN118105051A (en) | Rehabilitation cloud platform system for monitoring cardiopulmonary function | |
| CN112768074A (en) | Artificial intelligence-based serious disease risk prediction method and system | |
| WO2020237371A1 (en) | System and method for filtering time-varying data for physiological signal prediction | |
| CN114862865A (en) | Vessel segmentation method and system based on multi-view coronary angiography sequence image | |
| CN113096815A (en) | Chronic nephropathy prediction method based on logistic regression | |
| CN112884225A (en) | Body test result prediction method, body test result prediction device, electronic equipment and storage medium | |
| CN112733729A (en) | Model training and regression analysis method, device, storage medium and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |