CN114451879A

CN114451879A - Intelligent heart rate variability analysis system

Info

Publication number: CN114451879A
Application number: CN202210253155.0A
Authority: CN
Inventors: 朱佳兵; 何金蝉; 吕恒; 李毅
Original assignee: Wuhan Zoncare Bio Medical Electronics Co ltd
Current assignee: Wuhan Zoncare Bio Medical Electronics Co ltd
Priority date: 2022-03-15
Filing date: 2022-03-15
Publication date: 2022-05-10
Anticipated expiration: 2042-03-15
Also published as: CN114451879B

Abstract

The application discloses intelligent heart rate variability analytic system, this system includes: the device comprises a signal acquisition module, a preprocessing module, a correction module and a report output module; the signal acquisition module is used for acquiring an electrocardiogram digital signal; the preprocessing module is used for correcting the electrocardiogram digital signals to obtain electrocardiogram data; the correction module is used for processing the electrocardiogram data to obtain an NN interval sequence; and the report output module is used for obtaining a heart rate variability analysis report according to the NN interval sequence. The invention realizes the full utilization of the existing medical resources, can automatically improve the analysis algorithm of heart rate variability, improves the accuracy of the automatic analysis result of the heart rate variability, and can meet the use requirements of multiple scenes and individuation of users.

Description

Intelligent heart rate variability analysis system

Technical Field

The invention relates to the technical field of signal processing, in particular to an intelligent heart rate variability analysis system.

Background

The heart rate variability refers to the subtle time change and the law thereof among successive cardiac cycles, and is an important index for evaluating the sympathetic-recurrent nerve tension and the balance thereof of the autonomic nervous system. Since the first recognition of the clinical relevance of heart rate variability, there are continuing studies that demonstrate that heart rate variability has important clinical value in the risk assessment of cardiovascular diseases such as heart failure, stroke, post-myocardial infarction, post-heart transplantation, etc. The medical community has achieved a general consensus: the heart rate variability can be used as a risk prediction index of acute myocardial infarction and an early warning index of diabetic neuropathy.

Currently, the industry standard analysis method for heart rate variability is: firstly, acquiring electrocardiosignals of a subject to complete digital-to-analog conversion and artifact identification; secondly, identifying a QRS complex wave reference point in the electrocardiosignal to obtain an RR interval sequence and editing the RR interval sequence to obtain a sinus NN interval; and thirdly, performing time domain analysis on the NN interval to obtain a correlation analysis result of the heart rate variability. In the analysis of the heart rate variability, because the time domain and frequency domain analysis is easily affected by inaccurate identification of the NN interval sequence, the automatic analysis result of the heart rate variability is not accurate, and a clinician needs to spend a great deal of time in the RR interval editing stage to remove unqualified RR intervals through visual inspection and manual correction. The existing automatic heart rate variability analysis system does not fully utilize historical information such as accurate NN intervals, heart beat types and the like obtained by editing and correcting by existing clinicians, has the problems of low intelligent degree and low analysis efficiency, and cannot provide personalized intelligent service according to the use requirements of users.

Therefore, the main problems of the existing heart rate variability analysis system are that the heart rate variability automatic analysis system is low in intelligent degree, medical historical data is not fully utilized, so that the analysis result is inaccurate, and the use requirements of multiple scenes cannot be met. There is a great need to design a heart rate variability analysis system, which can fully utilize the existing medical resources, automatically improve the analysis algorithm of heart rate variability, improve the accuracy of the automatic analysis result, and meet the multi-scenario and personalized use requirements of users.

Disclosure of Invention

In view of the above, there is a need to provide an intelligent heart rate variability analysis system, so as to solve the problems in the prior art that the heart rate variability analysis system has a low intelligence degree, an inaccurate analysis result, and cannot meet the personalized use requirements of the user.

In order to solve the above problems, the present invention provides an intelligent heart rate variability analysis system, comprising:

the device comprises a signal acquisition module, a preprocessing module, a correction module and a report output module;

the signal acquisition module is used for acquiring an electrocardiogram digital signal;

the preprocessing module is used for preprocessing the electrocardiogram digital signals to obtain electrocardiogram data;

the correction module is used for analyzing the electrocardiogram data to obtain an NN interval sequence;

and the report output module is used for obtaining a heart rate variability analysis report according to the NN interval sequence.

Further, the preprocessing module comprises: the device comprises an outlier detection unit, a filter unit and an identification unit;

an outlier detection unit for determining abnormal data in the electrocardiogram digital signal;

the filter unit is used for filtering the artifact in the electrocardiogram digital signal;

and the QRS identification unit is used for identifying the QRS complex datum point in the electrocardiogram digital signal.

Further, the correction module comprises an automatic correction unit, a manual correction unit, a storage unit, a neural network training unit and a prediction unit;

an automatic correction unit, configured to eliminate non-sinus cardiac RR intervals of the electrocardiographic data, resulting in a first NN interval sequence;

the manual correction unit is used for providing an interactive interface for a user, so that the user can manually correct and label the R-wave position and type of the first NN interval sequence and the RR interval sequence;

the storage unit is used for storing the electrocardiogram data after the manual correction and marking of the R wave position and type and the RR interval sequence;

the neural network training unit is used for establishing an initial correction model and training the initial correction model by using the electrocardiogram data in the storage unit to obtain a completely trained correction model;

and the prediction unit is used for carrying out QRS identification and RR interval automatic correction on electrocardiogram data by using the fully trained correction model.

Further, the automatic correction unit comprises a screening module and a calculation module;

the screening module is used for screening RR intervals which meet a first judgment condition in an RR interval sequence of electrocardiogram data and judging whether the RR intervals meet a second judgment condition; if the second judgment condition is met, recording the position of the RR interval in the RR interval sequence, and removing the RR interval from the RR interval sequence to obtain a first RR interval sequence; calculating a heart rate according to the first RR interval sequence, and removing RR intervals with the heart rate meeting a third judgment condition to obtain a second RR interval sequence;

and the calculation module is used for processing the second RR interval sequence by using a preset algorithm to obtain an NN interval sequence.

Further, the neural network training unit comprises a selection module, an interactive training module and a self-training module;

the selection module is used for providing an interface for a user so that the user can select an interactive training mode and a self-training mode;

the interactive training module is used for enabling a user to autonomously select an initial correction model and a neural network training method for neural network training;

and the self-training module is used for training the neural network according to a preset neural network and a training method.

Further, the interactive training module comprises a setting module used for enabling a user to set a data set division ratio, select a neural network model and a super-parameter tuning method.

Further, the self-training module comprises a network searching module, an architecture optimization module and a performance evaluation module,

the network searching module is used for storing a basic neural network model structure;

the architecture optimization module is used for carrying out hyper-parameter setting on the basic neural network model by utilizing a preset architecture optimization method to obtain a candidate correction model;

and the performance evaluation module is used for carrying out performance evaluation on the candidate correction model with complete training and selecting the candidate correction model with the best performance as the heart rate analysis model with complete training.

Further, the network searching module comprises a basic operation unit, and the basic operation unit forms the basic neural network model structure.

Further, the architecture optimization method comprises a gradient descent-based micro-web searching method.

Further, the report output module comprises a time domain analysis unit, a frequency domain analysis unit and a nonlinear analysis unit;

the time domain analysis unit is used for carrying out long-range and short-range heart rate variability time domain analysis according to the NN interval sequence to obtain a heart rate variability time domain analysis report;

the frequency domain analysis unit is used for performing long-range and short-range spectrum analysis according to the NN interval sequence to obtain a heart rate variability frequency domain analysis report;

and the nonlinear analysis unit is used for calculating nonlinear parameters of the NN interval sequence to obtain a heart rate variability nonlinear parameter analysis report.

Compared with the prior art, the invention has the beneficial effects that: performing digital-to-analog conversion on the electrocardiogram original signal through a signal acquisition module to obtain an electrocardiogram digital signal; the electrocardiograph digital signals are preprocessed through a preprocessing module, such as filtering, and a data basis is provided for subsequent heart rate variability analysis; the existing medical resources are fully utilized through the correction module, so that the accuracy of automatic identification and correction of the analysis system is greatly improved, personalized function selection can be provided for a user in the using process, and the multi-scene and personalized use requirements of the user are met; the heart rate variability is analyzed from multiple angles through the report output module, so that the analysis result is more comprehensive and accurate. The invention realizes the full utilization of the existing medical resources, can automatically improve the analysis algorithm of heart rate variability, improves the accuracy of the automatic analysis result of the heart rate variability, and can meet the multi-scene and personalized use requirements of users.

Drawings

FIG. 1 is a schematic structural diagram of an embodiment of an intelligent heart rate variability analysis system according to the present invention;

FIG. 2 is a schematic structural diagram of an embodiment of a preprocessing module provided in the present invention;

FIG. 3 is a schematic structural diagram of a calibration module according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a working flow of an embodiment of a neural network unit provided in the present invention;

FIG. 5 is a schematic structural diagram of a self-training module according to an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of another embodiment of a self-training module according to the present invention;

FIG. 7 is a schematic structural diagram of a network search module according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a report output module according to an embodiment of the present invention;

fig. 9 is a schematic diagram of a result obtained by performing spectrum analysis on the source database nsr001 according to an embodiment of the frequency domain analysis unit provided in the present invention;

fig. 10 shows Poincar's result after analyzing the source database nsr001 data according to an embodiment of the non-linear analysis unit provided in the present invention.

Detailed Description

The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate preferred embodiments of the invention and together with the description, serve to explain the principles of the invention and not to limit the scope of the invention.

The present invention provides an intelligent heart rate variability analysis system, which is described in detail below.

As shown in fig. 1, an embodiment of the present invention provides an intelligent heart rate variability analysis system 100, which includes a signal acquisition module 101, a preprocessing module 102, a correction module 103, and a report output module 104;

the signal acquisition module 101 is used for acquiring an electrocardiogram digital signal;

the preprocessing module 102 is configured to preprocess the electrocardiogram digital signal to obtain electrocardiogram data;

the correction module 103 is configured to analyze the electrocardiogram data to obtain an NN interval sequence;

the report output module 104 is configured to obtain a heart rate variability analysis report according to the NN interval sequence.

Compared with the prior art, the intelligent heart rate variability analysis system provided by the embodiment performs digital-to-analog conversion on the electrocardiogram original signal through the signal acquisition module to obtain an electrocardiogram digital signal; the electrocardiograph digital signals are preprocessed through a preprocessing module, such as filtering, and a data basis is provided for subsequent heart rate variability analysis; the existing medical resources are fully utilized through the correction module, so that the accuracy of automatic identification and correction of the analysis system is greatly improved, personalized function selection can be provided for a user in the using process, and the multi-scene and personalized use requirements of the user are met; the heart rate variability is analyzed from multiple angles through the report output module, so that the analysis result is more comprehensive and accurate.

As a specific embodiment, the signal acquisition module 101 includes a dynamic electrocardiograph, which is mainly used for acquiring an original signal of an electrocardiogram and obtaining an electrocardiogram digital signal after digital-to-analog conversion.

As a preferred embodiment, as shown in fig. 2, the preprocessing module 102 includes: outlier detecting section 201, filter section 202, and identifying section 203;

the outlier detecting unit 201 is configured to determine abnormal data in the electrocardiogram digital signal;

the filter unit 202 is configured to filter artifacts in the electrocardiogram digital signal;

the QRS identification unit 203 is configured to identify a QRS complex fiducial point in the electrocardiogram digital signal.

As a specific embodiment, the outlier detecting unit 201 is mainly used to find out a point inconsistent with the characteristics of other ecg sample points. Since the electrocardiographic data is one-dimensional data, the outlier detecting unit 201 employs outlier detection based on a statistical method in consideration of the operational performance. Specifically, the process of outlier detection is: firstly, respectively calculating the mean value and the variance of all electrocardiogram samples, and then taking the value exceeding the range of [ u-3 sigma, u +3 sigma ] as an outlier; where u represents the mean of the electrocardiogram samples and σ represents the standard deviation of the electrocardiogram samples.

As a specific embodiment, the filter unit 202 is mainly used for filtering various types of artifacts in the electrocardiogram. The specific treatment process comprises the following steps: and an FIR high-pass filter is used for filtering baseline drift, a Notch filter is used for filtering power frequency interference, and an IIR low-pass filter is used for filtering high-frequency electromyographic interference.

As a preferred embodiment, as shown in fig. 3, the correction module 103 includes an automatic correction unit 301, a manual correction unit 302, a storage unit 303, a neural network training unit 304, and a prediction unit 305;

the automatic correction unit 301 is configured to eliminate non-sinus heartbeat RR intervals of the electrocardiographic data, so as to obtain a first NN interval sequence;

the manual correction unit 302 is configured to provide an interactive interface for a user, so that the user performs manual correction and labeling on the R-wave position and type of the first NN interval sequence and the RR interval sequence;

the storage unit 303 is configured to store the electrocardiogram data after the manual correction and labeling of the R-wave position and type and the RR interval sequence;

the neural network training unit 304 is configured to establish an initial calibration model, and train the initial calibration model by using the electrocardiogram data in the storage unit to obtain a calibration model with complete training;

the prediction unit 305 is configured to perform QRS identification and RR interval automatic correction on electrocardiogram data by using the fully trained correction model.

As a preferred embodiment, as shown in fig. 3, the automatic correction unit 301 includes a filtering module 3011 and a calculation module 3012;

the screening module 3011 is configured to screen an RR interval that meets a first determination condition in an RR interval sequence of electrocardiogram data, and determine whether the RR interval meets a second determination condition; if the second judgment condition is met, recording the position of the RR interval in the RR interval sequence, and removing the RR interval from the RR interval sequence to obtain a first RR interval sequence; calculating a heart rate according to the first RR interval sequence, and removing RR intervals with the heart rate meeting a third judgment condition to obtain a second RR interval sequence;

the calculating module 3012 is configured to process the second RR interval sequence by using a preset algorithm to obtain a first NN interval sequence.

Generally, the non-sinus cardiac signal sources mixed into the sinus RR interval sequence (also called NN interval sequence) are mainly: (a) RR intervals produced by ectopic beats followed by long RR intervals, such as premature beats followed by compensatory pauses; (b) long RR intervals due to missing detection of QRS complexes; (c) false QRS due to multiple detection of QRS complexes may cause an otherwise normal RR interval of sinus heart beats to be divided into two short RR intervals, resulting in an erroneous RR interval sequence. In the scheme, the screening module 3011 performs judgment for three times, and the calculating module 3012 performs data processing, so that the detection accuracy of the NN interval sequence is greatly improved.

As a specific example, the first determination condition is: RR interval >2 s; the second judgment condition is as follows: cardiac arrest conditions; the third judgment condition is as follows: RR intervals with a heart rate greater than 200bpm or less than 30 bpm. Under the preset conditions, the data correction process of the screening module 3011 is as follows: firstly, screening out data with RR interval of more than 2s in electrocardiogram data, and judging whether the condition of cardiac arrest is met; if yes, recording the position of the data meeting the condition, and removing the RR interval from the RR sequence; otherwise, no processing is performed. Secondly, the heart rate is calculated according to the RR intervals, and the RR intervals with the heart rate more than 200bpm or less than 30bpm are eliminated.

The preset algorithm of the calculation module 3012 is linear interpolation or cubic spline interpolation; and interpolating the RR intervals subjected to the elimination operation by the screening module 3011 to obtain the RR intervals of sinus heart beats used for heart rate variability analysis, and recording the RR intervals as NN interval sequences.

As a preferred embodiment, as shown in fig. 3, the neural network training unit 304 includes a selection module 3041, an interactive training module 3042, and a self-training module 3043;

the selection module 3041 is configured to provide an interface for a user, so that the user selects an interactive training mode and a self-training mode;

the interactive training module 3042 is configured to enable a user to autonomously select an initial calibration model and a neural network training method for neural network training;

the self-training module 3043 is configured to perform neural network training according to a preset neural network and a training method.

As a specific example, the selecting module 3041 may be used for the user to autonomously select whether to start the neural network training mode. If the user selects to start the neural network self-training, the system firstly counts the number of available label data in the designated storage unit, and in order to ensure the effectiveness of the neural network training, the minimum threshold value of the available label data is set to be 1 ten thousand. If the data volume is too small, prompting a user that the function cannot be started; and if the data volume meets the requirement, prompting the user to select to start the network interaction training module or the self-training module (the self-training module is started by default to perform neural network training). When a user has scientific research requirements, the training method experiment may need to be designed autonomously, for example, different models are set, relevant charts and analysis results are derived, and at this time, the interactive training module can be selected to be started.

As a preferred embodiment, the interactive training module 3042 includes a setting module, which is used for the user to set the data set partition ratio, select the neural network model and the super-parameter tuning method.

The neural network training unit 304 is described in detail below with reference to fig. 4.

As a specific embodiment, when the interactive training module 3042 is turned on, the training process is an interactive training mode, which specifically includes:

step S1: generating a visual exploration analysis report aiming at the existing data, comprising the following steps:

(a) descriptive statistics of the data, including conventional means, variances, medians, quantiles, etc.;

(b) data visualization charts, including histograms (visualization data distribution characteristics) of fields such as QRS wave position, QRS type, RR interval, etc., scatter charts (correlation between visualization characteristics), difference charts (difference describing distribution and statistical information in a plurality of data sets), box charts (visualization group difference), etc.;

step S2: setting a data set division ratio; two alternative division modes are available: training set-test set division, training set-verification set-test set;

step S3: selecting a model in a preset model library; common models such as a support vector machine, Logistic regression, random forest, CNN, FCN, ResNet, Unet and the like are provided in the model library for a user to select;

step S4: selecting a super parameter automatic tuning algorithm; and common algorithms such as Grid search and naive Bayesian evolution are provided for the user to select.

As a preferred embodiment, as shown in fig. 5, the self-training module 3043 includes a network search module 501, an architecture optimization module 502, and a performance evaluation module 503;

the network searching module 501 is configured to store a basic neural network model structure;

the architecture optimization module 502 performs hyper-parameter setting on the basic neural network model by using a preset architecture optimization method to obtain a candidate correction model;

the performance evaluation module 503 is configured to perform performance evaluation on the candidate calibration models that are completely trained, and select the calibration model with the best performance as the completely trained calibration model.

As a specific embodiment, as shown in fig. 6, the network searching module 501 also defines an algorithm for the user to search for the type of neural network. Meanwhile, in order to improve the search efficiency, the description mode of the neural network structure is customized. In this embodiment, the neural network structure in the network search module 501 is defined as a CNN and its variant network, an FCN and its variant network, a ResNet and its variant network, a Unet series network, and other structural networks.

The default of the architecture optimization method of the architecture optimization module 502 is a grid and random search mode, and the method enables a program to automatically run all parameters once through presetting a reasonable over-parameter value range, so as to obtain different performance index results. And finally, selecting a group of parameters with the optimal performance indexes as final values of the hyper-parameters. The architecture optimization module 502 can be utilized to perform, for each model in the web search module 501: and carrying out hyper-parameter setting on the model 1, the model 2 and the model … to obtain n candidate correction models.

After the training of each candidate correction model is completed, the performance evaluation module 503 uses the F1 score as a performance evaluation index, and selects the best model from the models 1, 2, and … n as the final completely trained correction model.

As a preferred embodiment, the network searching module 501 includes a basic operation unit, and the basic operation unit constitutes the basic neural network model structure.

As a specific embodiment, the network searching module 501 may be a hierarchical structure, that is, the unit structure generated in the previous step is used as a basic component of the unit structure in the next step, and a final network structure is obtained through iteration. As shown in fig. 7, fig. 7 is a schematic diagram of a three-layer structure. Combining the basic operation units of the first layer to be used as the basic operation units of the second layer; and then the main operation units of the second layer are combined to form a basic operation unit of a third layer.

As a preferred embodiment, the architecture optimization method comprises a gradient descent-based micro-web search method.

As a specific embodiment, darts (scalable Architecture search) is selected as the Architecture optimization method, which is a differentiable grid search method based on a gradient descent method, and searches for neural structures in a continuous and differentiable search space by using softmax to widen a discrete space.

As a specific example, the prediction unit 305 is activated when the user starts the neural network self-training module and completes the training of the neural network. After the prediction unit 305 is started, when the user subsequently uses the heart rate variability analysis function again, the prediction unit 305 will be used for the identification of QRS and the automatic correction of RR intervals.

As a preferred embodiment, as shown in fig. 8, the report output module 104 includes a time domain analysis unit 1041, a frequency domain analysis unit 1042, and a nonlinear analysis unit 1043;

the time domain analysis unit 1041 is configured to perform long-range and short-range heart rate variability time domain analysis according to the NN interval sequence to obtain a heart rate variability time domain analysis report;

the frequency domain analysis unit 1042 is configured to perform long-range and short-range spectrum analysis according to the NN interval sequence to obtain a heart rate variability frequency domain analysis report;

the nonlinear analysis unit 1043 is configured to calculate a nonlinear parameter of the NN interval sequence, and obtain a heart rate variability nonlinear parameter analysis report.

As a specific example, the time domain analysis of the time domain analysis unit 1041 is used for short-range (e.g. 5mins) heart rate variability detection and analysis, and more mainly for long-range (e.g. 24h) heart rate variability detection and analysis. The method comprises the following steps:

(1) NN interval histogram: and counting the NN interval distribution map in a certain time. Specifically, the number of heartbeats of different NN intervals is counted at a certain NN interval (e.g., 1/128s, i.e., 7.8124 ms). The abscissa is the length of the NN interval and the ordinate is the number of heart beats.

(2) NN interval difference histogram: a profile of the difference of adjacent NN intervals. Typically, the NN interval of the previous heart beat is subtracted from the NN interval of the next heart beat. The abscissa is the NN interval difference in ms and the ordinate is the number of heart beats.

(3) Day-night mean NN interval difference (ms): the NN intervals recorded at 24 hours are divided into two segments by day and night (sleeping time), the average value of the NN intervals at all nights is calculated, and the average value of the NN intervals at all daytime is subtracted to obtain the day-night average NN interval difference. Since the heart rate is slow during sleep at night and the NN interval is longer than during the day, the average NN interval difference between day and night is generally positive.

(4) NN interval Standard Deviation (SDNN): and the unit is ms, and the standard deviation of all NN intervals in a preset time period is detected.

(5) Standard deviation of mean of NN intervals (SDANN): the unit is ms, the NN interval data measured in 24h is divided into a plurality of segments (24h, 288 segments in total) by taking each 5min as a segment according to the time sequence, the average value of the NN intervals in each 5mins time segment is firstly calculated, the average value of 288 NN intervals can be obtained, and then the standard deviation of the 288 NN intervals is calculated.

(6) Root mean square of adjacent NN interval difference (RMSSD): the unit is ms. Adjacent NN intervals here mean that two heart beats are adjacent and at the same time both heart beats meet the criteria for analyzing the raw data as heart rate variability. For example, if both heart beats are sinus rhythms and there are no ectopic beats, missed heart beats, etc. after that, RMSSD may be expressed as:

wherein Δ NN_iRepresenting the difference between two adjacent NN intervals, there are N pairs of adjacent NN intervals in 24 h.

(7) Mean value of NN interval standard deviation (SDNN index): the unit is ms, the NN interval obtained by long-range detection is divided into every 5mins in time sequence, the standard deviation of the NN interval in each time period is firstly calculated, and then the average value of the standard deviations is calculated.

(8) Standard deviation of adjacent NN interval difference (SDSD): the unit is ms, the difference values of all adjacent NN intervals are calculated firstly, and then the standard deviation of the difference value data is calculated.

(9) Number of heart beats with difference between adjacent NN intervals exceeding 50ms (NN 50): the unit is heart beat number, and the index is generally used for long-range 24h heart rate variability analysis.

(10) The number of heart beats with a difference of more than 50ms between adjacent NN intervals as a percentage of the total number of NN intervals (pNN 50): and is substantially the same as NN 50. In contrast, when NN50 is used, the heart rate variability measurement time must be strictly defined to be 24 h.

(11) Heart rate variability trigonometric index (heart rate variability triangular index): the total number of heartbeats of the NN interval is divided by the height of the histogram of the NN interval (i.e. the number of heartbeats at the highest point of the NN interval histogram). For normalization, the NN interval histogram must typically be plotted at 7.8125ms (1/128s) intervals.

(12) Width of NN interval histogram (TRNN): in ms, it describes approximately the shape of the NN interval histogram with a triangle whose vertex is the highest point of the NN interval histogram.

(13) Heart rate variability difference index: the width of the histogram is calculated in ms at two set heights (e.g. 1000 heart beats and 10000 heart beats) of the adjacent NN interval difference histogram, and the difference between the two widths is the difference index.

(14) Logarithmic heart rate variability index: in the adjacent NN interval difference histogram, the abscissa is the absolute value of the difference of NN intervals of two adjacent heartbeats.

(15) Lorenz scattergrams: the NN interval NNn of the previous heart beat in the adjacent NN intervals is used as the abscissa, and the NN interval NNn +1 of the next heart beat is used as the ordinate. Plotting these points across the NN interval data obtained for long range detection yields a Lorenz scatterplot.

As a specific example, the frequency domain analysis of the frequency domain analyzing unit 1042 can be used for performing spectrum analysis for both short-range detection of 5min and long-range detection of 24 h. But it is recommended to use a short range detection of 5min for the spectral analysis. Short-range detection for 5min includes:

(1) total Power (TP): unit ms²An integral value of the power spectral density curve in the range of 0.0-0.4 Hz;

(2) ultra low frequency power (ULF): the unit ms2, is the integral value of the power spectral density curve in the range of 0.003-0.04 Hz;

(3) low frequency power (LF): in units of ms²LF is the integral value of the power spectral density curve in the range of 0.04-0.15 Hz;

(4) high frequency power (HF): in units of ms²HF is the integral value of the power spectral density curve in the range of 0.15-0.4 Hz;

(5) low frequency high frequency power ratio (LF/HF);

(6) normalized LF power (lfnu);

(7) normalized HF power (hfnu).

As shown in fig. 9, fig. 9 is a schematic diagram of a result obtained by performing spectrum analysis on the source database nsr001 by using the frequency domain analysis unit 1042.

As a specific example, the non-linearity analyzing unit 1043 mainly calculates some relevant poincare non-linearity parameters. The specific analysis method comprises the following steps:

(1) SD1 Standard deviation of Poincare figure projected onto a straight line perpendicular to the identity;

(2) SD 2: the Poincare diagram projection y is the standard deviation on the x straight line;

(3)SD2/SD1。

as shown in fig. 10, fig. 10 is a Poincar é result after analyzing nsr001 data of the source database by the nonlinear analysis unit 1043. Wherein SD1 is 22.1526, SD2 is 236.9698, and SD2/SD1 value is 10.6971.

The invention discloses an intelligent heart rate variability analysis system, which performs digital-to-analog conversion on an electrocardiogram original signal through a signal acquisition module to obtain an electrocardiogram digital signal; the electrocardiograph digital signals are preprocessed through a preprocessing module, such as filtering, and a data basis is provided for subsequent heart rate variability analysis; through the correction module, information corrected by a doctor can be fully utilized by utilizing a neural network learning mode, the accuracy of an algorithm is continuously improved, the existing medical resources are fully utilized, the accuracy of automatic identification and correction of an analysis system is greatly improved, personalized function selection can be provided for a user in the using process, and the multi-scene and personalized use requirements of the user are met; the heart rate variability is analyzed from multiple angles through the report output module, so that the analysis result is more comprehensive and accurate.

The invention realizes the full utilization of the existing medical resources, can automatically improve the analysis algorithm of heart rate variability, improves the accuracy of the automatic analysis result of the heart rate variability and achieves the complete automatic diagnosis of the heart rate variability; and can meet the multi-scene and personalized use requirements of users.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.

Claims

1. An intelligent heart rate variability analysis system, comprising: the device comprises a signal acquisition module, a preprocessing module, a correction module and a report output module;

2. The intelligent heart rate variability analysis system of claim 1, wherein the preprocessing module comprises: the device comprises an outlier detection unit, a filter unit and an identification unit;

3. The intelligent heart rate variability analysis system of claim 1, wherein the correction module comprises an automatic correction unit, a manual correction unit, a storage unit, a neural network training unit, and a prediction unit;

and the prediction unit is used for carrying out QRS identification and RR interval automatic correction on electrocardiogram data by using the correction model with complete training.

4. The intelligent heart rate variability analysis system of claim 3, wherein the automated correction unit comprises a screening module and a calculation module;

and the calculation module is used for processing the second RR interval sequence by using a preset algorithm to obtain a first NN interval sequence.

5. The intelligent heart rate variability analysis system of claim 3, wherein the neural network training unit comprises a selection module, an interactive training module, and a self-training module;

6. The intelligent heart rate variability analysis system of claim 5, wherein the interactive training module comprises a setup module for the user to set up data set partitioning ratios, select neural network models, and hyper-parametric tuning methods.

7. The intelligent heart rate variability analysis system of claim 5, wherein the self-training module comprises a web search module, an architecture optimization module, and a performance evaluation module,

and the performance evaluation module is used for carrying out performance evaluation on the candidate correction model with complete training and selecting the correction model with the best performance as the correction model with complete training.

8. The intelligent heart rate variability analysis system of claim 7, wherein the network search module comprises basic arithmetic units, and the basic arithmetic units form the basic neural network model structure.

9. The intelligent heart rate variability analysis system of claim 7 wherein the architecture optimization method comprises a gradient descent based micro-web search method.

10. The intelligent heart rate-variability analysis system of claim 1, wherein said report output module comprises a time domain analysis unit, a frequency domain analysis unit, and a non-linear analysis unit;