CN115067910A - Heart rate variability pressure detection method, device, storage medium and system - Google Patents
Heart rate variability pressure detection method, device, storage medium and system Download PDFInfo
- Publication number
- CN115067910A CN115067910A CN202210859450.0A CN202210859450A CN115067910A CN 115067910 A CN115067910 A CN 115067910A CN 202210859450 A CN202210859450 A CN 202210859450A CN 115067910 A CN115067910 A CN 115067910A
- Authority
- CN
- China
- Prior art keywords
- heart rate
- feature set
- rate variability
- feature
- preset
- 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.)
- Pending
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/024—Detecting, measuring or recording pulse rate or heart rate
- A61B5/02405—Determining heart rate variability
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Cardiology (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Signal Processing (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Power Engineering (AREA)
- Fuzzy Systems (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a heart rate variability pressure detection method, a heart rate variability pressure detection device, a storage medium and a heart rate variability pressure detection system. The electrocardiogram signal is filtered to obtain a first detection signal, heart rate data and heart beat interval data are extracted from the first detection signal, time domain, frequency domain and nonlinear three-aspect analysis is carried out on heart rate variability according to the heart beat interval data to obtain a heart rate variability feature set, feature dimensionality reduction is carried out on a feature set to be detected obtained by fusing the heart rate data and the heart rate variability feature set by using FCA-Relieff to obtain a dimensionality reduction feature set, and identification and classification are carried out according to a preset multilayer perceptron model.
Description
Technical Field
The invention relates to the technical field of heart rate variability pressure detection, in particular to a heart rate variability pressure detection method, a heart rate variability pressure detection device, a heart rate variability pressure detection storage medium and a heart rate variability pressure detection system.
Background
With the continuous improvement of the living standard of the substance, people pay more and more attention to the physical and psychological health of individuals. Among them, stress is one of the most important mental health problems, and a long-term stress condition may cause headache, cardiovascular diseases, depression, etc., thereby seriously damaging physical health and mental health. There would be significant positive social and personal benefits if stress could be detected continuously at an early stage. Existing stress detection is based on different detection mechanisms, such as video-based stress detection, voice-based stress detection, and stress detection based on physiological signals. Studies have shown that physiological signals are a reliable indicator of predicted stress. Physiological signals can be obtained by various sensors.
In the prior art, most of the pressure detection methods are based on the combination of physiological signals, including signals collected from sensors such as electrocardiogram, electrodermal activity and electromyogram, and then the physiological signals are analyzed by a machine learning algorithm to detect the pressure. For example, patent No. CN114139571A discloses a pressure detection method and system based on multi-modal physiological signals, the multi-modal physiological signals of the method include skin electrical activity signals, photoplethysmography signals, and skin temperature signals, the physiological signals are preprocessed and then transmitted into a pressure detection migration model, the model includes a feature extractor and a pressure classifier, the feature extractor is used for extracting depth features and fusing the depth features with artificial time-frequency domain features, and the pressure classifier is used for classifying pressure levels of users.
However, the prior art still has the following defects: the prior art is mainly based on the detection of multi-modal physiological signals, which increases the complexity of detection: the observed person needs to wear the wearable device, and although the wearable device is a non-invasive device, the wearable device has certain influence on the normal activities of people; meanwhile, the detection of various physiological signals has a certain delay, so that the various signals have time deviation, and the accuracy of detection and judgment is low.
Therefore, there is a need for a heart rate variability pressure detection method, apparatus, storage medium, and system that overcome the above-mentioned deficiencies in the prior art.
Disclosure of Invention
The embodiment of the invention provides a heart rate variability pressure detection method, a heart rate variability pressure detection device, a storage medium and a heart rate variability pressure detection system, so that the accuracy of detection judgment is improved.
An embodiment of the present invention provides a heart rate variability pressure detection method, including: acquiring an electrocardiogram signal of an object to be detected, and filtering the electrocardiogram signal to acquire a first detection signal; extracting heart rate data and heartbeat interval data from the first detection signal, and performing time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data to obtain a heart rate variability feature set; fusing the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and performing dimension reduction processing on the feature set to be detected according to a preset FCA-Relieff dimension reduction method to obtain a dimension reduction feature set; and according to a preset multilayer perceptron model, identifying and classifying the dimension reduction characteristic set to obtain a detection result.
As an improvement of the above scheme, the filtering the electrocardiogram signal to obtain the first detection signal specifically includes: rejecting outliers in the electrocardiogram signal to obtain a first electrocardiogram signal; and according to a preset FIR band-pass filtering method and a preset band-pass frequency range, filtering the first electrocardiogram signal to obtain a first detection signal.
As an improvement of the above scheme, time domain analysis, frequency domain analysis and nonlinear analysis are performed according to the heartbeat interval data, and a heart rate variability feature set is obtained, specifically including: performing overall variability analysis on the heartbeat interval data according to a preset HRV function to obtain a time domain feature set; according to a preset HRV function, carrying out specific frequency band analysis on the heartbeat interval data to obtain a frequency domain characteristic set; analyzing the cardiovascular system regulation effect of the heartbeat interval data according to a preset HRV function to obtain a nonlinear feature set; and acquiring a heart rate variability feature set according to the time domain feature set, the frequency domain feature set and the nonlinear feature set.
As an improvement of the above scheme, according to a preset FCA-ReliefF dimension reduction method, the dimension reduction processing is performed on the feature set to be detected to obtain a dimension reduction feature set, which specifically includes: calculating the correlation among the characteristics to be detected in the characteristic set to be detected, and deleting redundant characteristics from the characteristic set to be detected according to the correlation and a preset redundant threshold value to obtain an effective characteristic set; and performing dimension reduction screening processing on the effective feature set according to a preset Relieff algorithm and a preset dimension reduction screening threshold value to obtain a dimension reduction feature set.
As an improvement of the above scheme, according to the correlation and a preset redundancy threshold, deleting redundant features from the feature set to be detected to obtain an effective feature set, specifically including: judging whether the correlation is greater than a preset redundancy threshold value or not; if so, acquiring a first relevant feature and a second relevant feature corresponding to the correlation, respectively testing to obtain a first model influence factor of the first relevant feature and a second model influence factor of the second relevant feature, and judging the magnitude relation of the first model influence factor and the second model influence factor; when the first model influence factor is larger than the second model influence factor, taking the second relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set; and when the first model influence factor is smaller than the second model influence factor, taking the first relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set.
As an improvement of the above, after extracting heart rate data and heartbeat interval data from the first detection signal, the detection method further includes: and denoising the heartbeat interval data and the heart rate data.
As an improvement of the above scheme, the denoising processing is performed on the heartbeat interval data and the heart rate data, and specifically includes: and carrying out median replacement and median filtering on the heartbeat interval data, and carrying out mean replacement and median filtering on the heart rate data.
The invention correspondingly provides a heart rate variability pressure detection device, which comprises a signal acquisition unit, a multi-dimensional analysis unit, a fusion dimensionality reduction unit and a classification detection unit, wherein the signal acquisition unit is used for acquiring an electrocardiogram signal of an object to be detected and filtering the electrocardiogram signal to acquire a first detection signal; the multi-dimensional analysis unit is used for extracting heart rate data and heartbeat interval data from the first detection signal, and performing time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data to obtain a heart rate variability feature set; the fusion dimensionality reduction unit is used for carrying out fusion processing on the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and carrying out dimensionality reduction processing on the feature set to be detected according to a preset FCA-Relieff dimensionality reduction method to obtain a dimensionality reduction feature set; and the classification detection unit is used for identifying and classifying the dimension reduction characteristic set according to a preset multilayer perceptron model so as to obtain a detection result.
As an improvement of the above, the signal acquisition unit is further configured to: rejecting outliers in the electrocardiogram signal to obtain a first electrocardiogram signal; and according to a preset FIR band-pass filtering method and a preset band-pass frequency range, filtering the first electrocardiogram signal to obtain a first detection signal.
As an improvement of the above, the multidimensional analysis unit is further configured to: performing overall variability analysis on the heartbeat interval data according to a preset HRV function to obtain a time domain feature set; according to a preset HRV function, carrying out specific frequency band analysis on the heartbeat interval data to obtain a frequency domain characteristic set; analyzing the cardiovascular system regulation effect of the heartbeat interval data according to a preset HRV function to obtain a nonlinear feature set; and acquiring a heart rate variability feature set according to the time domain feature set, the frequency domain feature set and the nonlinear feature set.
As an improvement of the above solution, the fusion dimension reduction unit is further configured to: calculating the correlation among the characteristics to be detected in the characteristic set to be detected, and deleting redundant characteristics from the characteristic set to be detected according to the correlation and a preset redundant threshold value to obtain an effective characteristic set; and performing dimension reduction screening processing on the effective feature set according to a preset Relieff algorithm and a preset dimension reduction screening threshold value to obtain a dimension reduction feature set.
As an improvement of the above solution, the fusion dimension reduction unit is further configured to: judging whether the correlation is larger than a preset redundancy threshold value or not; if so, acquiring a first relevant feature and a second relevant feature corresponding to the correlation, respectively testing to obtain a first model influence factor of the first relevant feature and a second model influence factor of the second relevant feature, and judging the magnitude relation of the first model influence factor and the second model influence factor; when the first model influence factor is larger than the second model influence factor, taking the second relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set; and when the first model influence factor is smaller than the second model influence factor, taking the first relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set.
As an improvement of the above solution, the detection apparatus further includes a denoising processing unit, and the denoising processing unit is configured to: and denoising the heartbeat interval data and the heart rate data.
As an improvement of the above solution, the denoising processing unit is further configured to: and carrying out median replacement and median filtering on the heartbeat interval data, and carrying out mean replacement and median filtering on the heart rate data.
Another embodiment of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed, controls a device on which the computer-readable storage medium is located to perform the heart rate variability pressure detection method as described above.
Another embodiment of the invention provides a heart rate variability pressure detection system comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the heart rate variability pressure detection method as described above when executing the computer program.
Compared with the prior art, the technical scheme has the following beneficial effects:
the invention provides a heart rate variability pressure detection method, a heart rate variability pressure detection device, a computer readable storage medium and a heart rate variability pressure detection system.
Drawings
Fig. 1 is a schematic flow chart of a heart rate variability pressure detection method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a heart rate variability pressure detection device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Detailed description of the preferred embodiment
The embodiment of the invention firstly describes a heart rate variability pressure detection method. Fig. 1 is a schematic flow chart of a heart rate variability pressure detection method according to an embodiment of the present invention.
As shown in fig. 1, the detection method includes:
s1, obtaining an electrocardiogram signal of the object to be detected, and filtering the electrocardiogram signal to obtain a first detection signal.
Either the electrocardiogram signal or the PPG signal, here exemplified by the electrocardiogram signal, may be used to obtain heart rate variability data. And (3) performing noise filtering on the electrocardiogram signals by adopting a BioSPPy tool kit, wherein the noise filtering comprises elimination of abnormal values and FIR band-pass filtering. The FIR band-pass filtering can limit the frequency of the waves needing to be reserved within a certain range, is used for removing surrounding noise and can achieve a good effect. The electrocardiogram signal waveform after FIR band-pass filtering is smoother.
In one embodiment, filtering the electrocardiogram signal to obtain the first detection signal includes: rejecting outliers in the electrocardiogram signal to obtain a first electrocardiogram signal; and according to a preset FIR band-pass filtering method and a preset band-pass frequency range, filtering the first electrocardiogram signal to obtain a first detection signal.
And S2, extracting heart rate data and heartbeat interval data from the first detection signal, and performing time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data to obtain a heart rate variability feature set.
After the first detection signal is obtained, a biossppy tool kit is also adopted to extract heart rate and RR interval information of the electrocardiogram signal, wherein the RR interval is an interval between a peak value R and an R point in an electrocardiogram signal wave group, and the QRS wave is identified based on a biospy. The Heart Rate (HR) is calculated by dividing 1 by the RR interval RRI in seconds. Because the pressure sensor can be extracted from electrocardiogram PPG signals and imaging type photoplethysmography IPPG signals, the IPPG technology can realize non-contact pressure detection, and the detection mode is simpler and more convenient and is easy to use in a large scale.
The heart rate variability calculation including time domain analysis, frequency domain analysis and nonlinear analysis was performed on the RR intervals using the hrv function in the neurolit 2 toolkit, resulting in 72 features in total. The time domain analysis method is to carry out statistical analysis on R-R interval signals, and reflects the overall variability of heart rate; the frequency domain analysis may reflect the respective excitations of the sympathetic and parasympathetic nerves, the frequency indices of the power distributions of the different frequency bands being particularly suitable for assessing specific components of the heart rate variability, the high and low frequency bands being significant reflections of the parasympathetic and sympathetic nerve activities, respectively; the non-linear analysis method can be used to evaluate the modulation of the cardiovascular system, and the HRV features obtained from the non-linear analysis method can fully characterize the dynamics of the heart. Setting a sliding window to be 20s, respectively adding a part of samples from the previous window and the next window, wherein the data of the samples are determined according to the sampling frequency, and the actual length of the window is about 30 s. The use of overlapping windows avoids hard cuts. Such a method may include samples at the beginning or end of a window to calculate more RR peaks in calculating the heart rate variability features. The following table demonstrates the common temporal, frequency domain and nonlinear characteristics of heart rate variability.
In one embodiment, performing time domain analysis, frequency domain analysis and nonlinear analysis on the heartbeat interval data to obtain a heart rate variability feature set specifically includes: performing overall variability analysis on the heartbeat interval data according to a preset HRV function to obtain a time domain feature set; according to a preset HRV function, carrying out specific frequency band analysis on the heartbeat interval data to obtain a frequency domain characteristic set; analyzing the cardiovascular system regulation effect of the heartbeat interval data according to a preset HRV function to obtain a nonlinear feature set; and acquiring a heart rate variability feature set according to the time domain feature set, the frequency domain feature set and the nonlinear feature set.
Under the influence of noise, the RR interval data has some abnormal values which are obviously beyond the normal RR interval data range, the abnormal values are replaced by NA, and then the numerical values and the missing values of the RR interval which is less than 0.5 second or more than 1.5 seconds are replaced by a median so as to ensure that the abnormal values do not exceed the normal range of the RR interval. And performing median filtering after replacement, wherein the basic principle of the median filtering is to replace the value of one point in the data by the median of each point value in a neighborhood of the point so as to eliminate isolated noise, and smooth the signal, and the kernel number is 5. Firstly, replacing an abnormal value which is obviously out of range with NA, then replacing the NA value with an average value, and finally, further removing the unnatural heart rate by adopting a median filter with 13 kernels.
In one embodiment, after extracting heart rate data and heartbeat interval data from the first detection signal, the detection method further comprises: and denoising the heartbeat interval data and the heart rate data.
In an embodiment, denoising the heartbeat interval data and the heart rate data specifically includes: and carrying out median replacement and median filtering on the heartbeat interval data, and carrying out mean replacement and median filtering on the heart rate data.
And S3, fusing the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and performing dimension reduction processing on the feature set to be detected according to a preset FCA-Relieff dimension reduction method to obtain a dimension reduction feature set.
Fusing 72 features of heart rate variability with the heart rate signal for a total of 73 features; and because the characteristic dimension is higher, the dimension reduction is carried out by adopting a preset FCA-Relieff dimension reduction method. The method comprises the following steps: the first step is feature correlation analysis FCA, the correlation among all features is calculated, and a threshold value is set for deletion. If the correlation is greater than the threshold, it indicates that there is a high correlation between the two features, i.e., repeatability, and therefore one of the features needs to be deleted. The specific method for selecting the features to be deleted is to compare which feature is deleted, so that the model efficiency can be improved greatly. Specifically, one of the features is removed from the original data and the model result is tested; and removing another characteristic test model result from the original data. The decision as to which relevant feature to eventually delete is targeted at a higher model efficiency. Redundant features are removed, meanwhile, the capability of expressing information by the remaining features cannot be reduced, and the threshold value is generally set to be 0.85-0.95; and the second step is a Relieff algorithm which is an improved algorithm based on the Relief, a weight value is distributed to each feature according to the relevance degree between the feature and the category to which the feature belongs, a threshold value is set for screening, the Relief is only limited to a two-classification problem, and the improved Relieff is suitable for a multi-classification problem and can be applied to multi-classification pressure model feature selection. The thresholds of both FCA and ReliefF analysis methods may be explored by setting a threshold gradient to obtain the optimal threshold.
In one embodiment, according to a preset FCA-ReliefF dimension reduction method, the dimension reduction processing is performed on the feature set to be detected to obtain a dimension reduction feature set, which specifically includes: calculating the correlation among the characteristics to be detected in the characteristic set to be detected, and deleting redundant characteristics from the characteristic set to be detected according to the correlation and a preset redundant threshold value to obtain an effective characteristic set; and performing dimension reduction screening processing on the effective feature set according to a preset Relieff algorithm and a preset dimension reduction screening threshold value to obtain a dimension reduction feature set.
In one embodiment, deleting redundant features from the feature set to be detected to obtain an effective feature set according to the correlation and a preset redundant threshold specifically includes: judging whether the correlation is larger than a preset redundancy threshold value or not; if so, acquiring a first relevant feature and a second relevant feature corresponding to the correlation, respectively testing to obtain a first model influence factor of the first relevant feature and a second model influence factor of the second relevant feature, and judging the magnitude relation of the first model influence factor and the second model influence factor; when the first model influence factor is larger than the second model influence factor, taking the second relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set; and when the first model influence factor is smaller than the second model influence factor, taking the first relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set.
S4: and according to a preset multilayer perceptron model, identifying and classifying the dimension reduction characteristic set to obtain a detection result.
A multi-layered perceptron Model (MLP) is capable of describing complex mapping relationships between a set of input variables to output variables. Taking the characteristic subjected to dimensionality reduction in the step S6 as an input layer of the model, wherein the number of neurons of the input layer is the same as the characteristic dimensionality; the number of hidden layers is generally multilayer, theoretically, a neural network of only one hidden layer can approach any continuous function as long as the neural network contains enough neurons, and research shows that the hidden layer of the MLP is preferably set to be 1-2 layers, and the number of the neurons of the hidden layer can be set with gradients for exploration to find the optimal number of the neurons; the number of neurons in the output layer is the number of output variables, if the pressure two classification is performed, the number of neurons in the output layer is 1, which represents the probability of belonging to the positive class, and if the pressure three classification is performed, the number of neurons in the output layer is 3, which represents the probability of belonging to a certain class.
The embodiment of the invention integrates the time domain, frequency domain and nonlinear characteristics of heart rate variability, FCA-Relieff characteristic dimension reduction and a multilayer perceptron model MLP. The heart rate variability can be extracted from electrocardiogram (PPG) signals and imaging type photoplethysmography (IPPG) signals, the IPPG technology can realize non-contact pressure detection, and the detection mode is simpler and more convenient and is easy to use in a large scale; analysis of heart rate variability is based on three aspects: 72-dimensional features can be extracted in all of time domain analysis, frequency domain analysis and nonlinear analysis, so that enough information can be obtained to predict pressure more accurately; feature dimensionality reduction is carried out by fusing feature correlation analysis FCA and Relieff algorithms, redundant features and features with low correlation with target values are removed, and the efficiency and accuracy of pressure identification can be improved; the multilayer perceptron model MLP obtains an optimal MLP model by performing gradient exploration on the number of layers of hidden layers and the number of neurons.
The embodiment of the invention discloses a heart rate variability pressure detection method, which comprises the steps of filtering an electrocardiogram signal to obtain a first detection signal, extracting heart rate data and heart rate interval data from the first detection signal, analyzing heart rate variability in three aspects of time domain, frequency domain and nonlinearity according to the heart rate interval data to obtain a heart rate variability feature set, using FCA-Relieff to perform feature dimensionality reduction on a feature set to be detected obtained by fusing the heart rate data and the heart rate variability feature set to obtain a dimensionality reduction feature set, and performing identification and classification according to a preset multilayer perceptron model.
Detailed description of the invention
Besides the method, the embodiment of the invention also discloses a heart rate variability pressure detection device. Fig. 2 is a schematic structural diagram of a heart rate variability pressure detection device according to an embodiment of the present invention.
As shown in fig. 2, the detection apparatus includes a signal acquisition unit 11, a multi-dimensional analysis unit 12, a fusion dimension reduction unit 13, and a classification detection unit 14.
The signal acquiring unit 11 is configured to acquire an electrocardiogram signal of an object to be detected, and filter the electrocardiogram signal to acquire a first detection signal.
In one embodiment, the signal obtaining unit 11 is further configured to: rejecting outliers in the electrocardiogram signal to obtain a first electrocardiogram signal; and according to a preset FIR band-pass filtering method and a preset band-pass frequency range, filtering the first electrocardiogram signal to obtain a first detection signal.
The multidimensional analysis unit 12 is configured to extract heart rate data and heartbeat interval data from the first detection signal, perform time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data, and acquire a heart rate variability feature set.
In one embodiment, the multidimensional analysis unit 12 is further configured to: performing overall variability analysis on the heartbeat interval data according to a preset HRV function to obtain a time domain feature set; according to a preset HRV function, carrying out specific frequency band analysis on the heartbeat interval data to obtain a frequency domain characteristic set; analyzing the cardiovascular system regulation function of the heartbeat interval data according to a preset HRV function to obtain a nonlinear feature set; and acquiring a heart rate variability feature set according to the time domain feature set, the frequency domain feature set and the nonlinear feature set.
The fusion dimensionality reduction unit 13 is configured to perform fusion processing on the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and perform dimensionality reduction processing on the feature set to be detected according to a preset FCA-ReliefF dimensionality reduction method to obtain a dimensionality reduction feature set.
In one embodiment, the fusion dimensionality reduction unit 13 is further configured to: calculating the correlation among the characteristics to be detected in the characteristic set to be detected, and deleting redundant characteristics from the characteristic set to be detected according to the correlation and a preset redundant threshold value to obtain an effective characteristic set; and performing dimension reduction screening processing on the effective feature set according to a preset Relieff algorithm and a preset dimension reduction screening threshold value to obtain a dimension reduction feature set.
In one embodiment, the fusion dimension reduction unit 13 is further configured to: judging whether the correlation is larger than a preset redundancy threshold value or not; if so, acquiring a first relevant feature and a second relevant feature corresponding to the correlation, respectively testing to obtain a first model influence factor of the first relevant feature and a second model influence factor of the second relevant feature, and judging the magnitude relation of the first model influence factor and the second model influence factor; when the first model influence factor is larger than the second model influence factor, taking the second relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set; and when the first model influence factor is smaller than the second model influence factor, taking the first relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set.
The classification detection unit 14 is configured to perform identification and classification on the dimension reduction feature set according to a preset multilayer perceptron model to obtain a detection result.
In one embodiment, the detection apparatus further comprises a denoising processing unit configured to: and denoising the heartbeat interval data and the heart rate data.
In one embodiment, the denoising processing unit is further configured to: and carrying out median replacement and median filtering on the heartbeat interval data, and carrying out mean replacement and median filtering on the heart rate data.
Wherein, the unit integrated with the detection device can be stored in a computer readable storage medium if the unit is realized in the form of a software functional unit and sold or used as an independent product. Based on such understanding, all or part of the flow in the method according to the above embodiments may be implemented by a computer program, which may be stored in a computer readable storage medium and used by a processor to implement the steps of the above embodiments of the method. That is, another embodiment of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed, controls a device on which the computer-readable storage medium is located to execute the heart rate variability pressure detection method as described above.
Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
It should be noted that the above-described device embodiments are merely illustrative, where the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the embodiment of the device provided by the invention, the connection relationship between the units indicates that the units have communication connection therebetween, and the connection relationship can be specifically implemented as one or more communication buses or signal lines. One of ordinary skill in the art can understand and implement it without inventive effort.
The embodiment of the invention discloses a heart rate variability pressure detection device and a computer readable storage medium, wherein a first detection signal is obtained by filtering an electrocardiogram signal, heart rate data and heart rate interval data are extracted from the first detection signal, heart rate variability is analyzed in three aspects of time domain, frequency domain and nonlinearity according to the heart rate interval data to obtain a heart rate variability feature set, FCA-Relieff is used for performing feature dimensionality reduction on a to-be-detected feature set obtained by fusing the heart rate data and the heart rate variability feature set to obtain a dimensionality reduction feature set, and recognition and classification are performed according to a preset multilayer sensor model.
Detailed description of the preferred embodiment
In addition to the above methods and apparatus, embodiments of the present invention also describe a heart rate variability pressure detection system.
The detection system comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the heart rate variability pressure detection method as described above when executing the computer program.
The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like that is the control center for the device and that connects the various parts of the overall device using various interfaces and lines.
The memory may be used to store the computer programs and/or modules, and the processor may implement the various functions of the apparatus by running or executing the computer programs and/or modules stored in the memory, as well as by invoking data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
The embodiment of the invention discloses a heart rate variability pressure detection system, which is used for filtering an electrocardiogram signal to obtain a first detection signal, extracting heart rate data and heart rate interval data from the first detection signal, analyzing heart rate variability in three aspects of time domain, frequency domain and nonlinearity according to the heart rate interval data to obtain a heart rate variability feature set, performing feature dimensionality reduction on a feature set to be detected obtained by fusing the heart rate data and the heart rate variability feature set by using FCA-Relieff to obtain a dimensionality reduction feature set, and performing identification and classification according to a preset multilayer perceptron model.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (10)
1. A heart rate variability stress detection method, the detection method comprising:
acquiring an electrocardiogram signal of an object to be detected, and filtering the electrocardiogram signal to acquire a first detection signal;
extracting heart rate data and heartbeat interval data from the first detection signal, and performing time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data to obtain a heart rate variability feature set;
fusing the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and performing dimension reduction processing on the feature set to be detected according to a preset FCA-Relieff dimension reduction method to obtain a dimension reduction feature set;
and according to a preset multilayer perceptron model, identifying and classifying the dimension reduction characteristic set to obtain a detection result.
2. The heart rate variability pressure detection method according to claim 1, wherein filtering the electrocardiogram signal to obtain a first detection signal, comprises:
rejecting outliers in the electrocardiogram signal to obtain a first electrocardiogram signal;
and according to a preset FIR band-pass filtering method and a preset band-pass frequency range, filtering the first electrocardiogram signal to obtain a first detection signal.
3. The heart rate variability pressure detection method according to claim 2, wherein said obtaining a heart rate variability feature set by performing a time domain analysis, a frequency domain analysis and a nonlinear analysis on said heartbeat interval data comprises:
performing overall variability analysis on the heartbeat interval data according to a preset HRV function to obtain a time domain feature set;
according to a preset HRV function, carrying out specific frequency band analysis on the heartbeat interval data to obtain a frequency domain characteristic set;
analyzing the cardiovascular system regulation effect of the heartbeat interval data according to a preset HRV function to obtain a nonlinear feature set;
and acquiring a heart rate variability feature set according to the time domain feature set, the frequency domain feature set and the nonlinear feature set.
4. The heart rate variability pressure detection method according to claim 3, wherein the dimension reduction processing is performed on the feature set to be detected according to a preset FCA-Relieff dimension reduction method to obtain a dimension reduction feature set, and specifically includes:
calculating the correlation among the characteristics to be detected in the characteristic set to be detected, and deleting redundant characteristics from the characteristic set to be detected according to the correlation and a preset redundant threshold value to obtain an effective characteristic set;
and performing dimension reduction screening processing on the effective feature set according to a preset Relieff algorithm and a preset dimension reduction screening threshold value to obtain a dimension reduction feature set.
5. The heart rate variability pressure detection method according to claim 4, wherein based on the correlation and a preset redundancy threshold, deleting redundant features from the set of features to be tested to obtain a set of valid features, specifically comprising:
judging whether the correlation is larger than a preset redundancy threshold value or not;
if so, acquiring a first relevant feature and a second relevant feature corresponding to the correlation, respectively testing to obtain a first model influence factor of the first relevant feature and a second model influence factor of the second relevant feature, and judging the magnitude relation of the first model influence factor and the second model influence factor;
when the first model influence factor is larger than the second model influence factor, taking the second relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set;
and when the first model influence factor is smaller than the second model influence factor, taking the first relevant feature as a redundant feature, and deleting the redundant feature from the feature set to be detected to obtain an effective feature set.
6. The heart rate variability pressure detection device according to any one of claims 1-5 wherein after extracting heart rate data and heartbeat interval data from the first detection signal, the detection method further comprises:
and denoising the heartbeat interval data and the heart rate data.
7. The heart rate variability pressure detecting device according to claim 6, wherein de-noising the heartbeat interval data and the heart rate data comprises:
and carrying out median replacement and median filtering on the heartbeat interval data, and carrying out mean replacement and median filtering on the heart rate data.
8. A heart rate variability pressure detection device is characterized by comprising a signal acquisition unit, a multi-dimensional analysis unit, a fusion dimension reduction unit and a classification detection unit,
the signal acquisition unit is used for acquiring an electrocardiogram signal of an object to be detected and filtering the electrocardiogram signal to acquire a first detection signal;
the multi-dimensional analysis unit is used for extracting heart rate data and heartbeat interval data from the first detection signal, and performing time domain analysis, frequency domain analysis and nonlinear analysis according to the heartbeat interval data to obtain a heart rate variability feature set;
the fusion dimensionality reduction unit is used for carrying out fusion processing on the heart rate variability feature set and the heart rate data to obtain a feature set to be detected, and carrying out dimensionality reduction processing on the feature set to be detected according to a preset FCA-Relieff dimensionality reduction method to obtain a dimensionality reduction feature set;
and the classification detection unit is used for identifying and classifying the dimension reduction characteristic set according to a preset multilayer perceptron model so as to obtain a detection result.
9. A computer-readable storage medium, comprising a stored computer program, wherein the computer-readable storage medium, when executed, controls an apparatus in which the computer-readable storage medium is located to perform a heart rate variability pressure detection method according to any one of claims 1 to 7.
10. Heart rate variability pressure detection system, characterized in that the detection system comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, when executing the computer program, implementing the heart rate variability pressure detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210859450.0A CN115067910A (en) | 2022-07-21 | 2022-07-21 | Heart rate variability pressure detection method, device, storage medium and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210859450.0A CN115067910A (en) | 2022-07-21 | 2022-07-21 | Heart rate variability pressure detection method, device, storage medium and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115067910A true CN115067910A (en) | 2022-09-20 |
Family
ID=83241926
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210859450.0A Pending CN115067910A (en) | 2022-07-21 | 2022-07-21 | Heart rate variability pressure detection method, device, storage medium and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115067910A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117271977A (en) * | 2023-09-27 | 2023-12-22 | 北京津发科技股份有限公司 | HRV data preprocessing method and device and electronic equipment |
-
2022
- 2022-07-21 CN CN202210859450.0A patent/CN115067910A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117271977A (en) * | 2023-09-27 | 2023-12-22 | 北京津发科技股份有限公司 | HRV data preprocessing method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107951485B (en) | Ambulatory ECG analysis method and apparatus based on artificial intelligence self study | |
| US20210000426A1 (en) | Classification system of epileptic eeg signals based on non-linear dynamics features | |
| EP3692903A1 (en) | Artificial intelligence self-learning-based automatic electrocardiography analysis method and apparatus | |
| CN108090509B (en) | Data length self-adaptive electrocardiogram classification method | |
| EP3698708A1 (en) | Ecg information processing method and ecg workstation | |
| Satheeskumaran et al. | Real-time ECG signal pre-processing and neuro fuzzy-based CHD risk prediction | |
| Liu et al. | Survey on atrial fibrillation detection from a single-lead ECG wave for Internet of Medical Things | |
| Sowmya et al. | Contemplate on ECG signals and classification of arrhythmia signals using CNN-LSTM deep learning model | |
| US20230352180A1 (en) | Atrial fibrillation risk prediction system based on heartbeat rhythm signals and application thereof | |
| WO2022135449A1 (en) | Interictal epileptiform discharge activity detection apparatus and method for epileptic patient | |
| KR102482796B1 (en) | Device and method for non-contact hrv analysis using remote ppg | |
| Rangappa et al. | Classification of cardiac arrhythmia stages using hybrid features extraction with k-nearest neighbour classifier of ecg signals | |
| Zhang et al. | Hybrid feature fusion for classification optimization of short ECG segment in IoT based intelligent healthcare system | |
| Yang et al. | Automated intra-patient and inter-patient coronary artery disease and congestive heart failure detection using EFAP-Net | |
| Abbaszadeh et al. | Probabilistic prediction of Epileptic Seizures using SVM | |
| CN115089179A (en) | Psychological emotion insights analysis method and system | |
| CN115067910A (en) | Heart rate variability pressure detection method, device, storage medium and system | |
| CN109674474B (en) | Sleep apnea recognition method, device and computer readable medium | |
| Parmar et al. | ECG signal based automated hypertension detection using fourier decomposition method and cosine modulated filter banks | |
| CA3065208A1 (en) | Long qt syndrome diagnosis and classification | |
| Sriraam et al. | Multichannel EEG based inter-ictal seizures detection using Teager energy with backpropagation neural network classifier | |
| Al-hajjar et al. | Epileptic seizure detection using feature importance and ML classifiers | |
| RU2751137C1 (en) | Method for determining sleep phase in long-term eeg recording | |
| CN113425298A (en) | Method for analyzing depression degree by collecting data through wearable equipment | |
| Seo et al. | A Deep Neural Network Based Wake-After-Sleep-Onset Time Aware Sleep Apnea Severity Estimation Scheme Using Single-Lead ECG Data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |