A fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities

Abstract

The present invention relates to the field of data processing technology, and specifically to a fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities, comprising: a data acquisition module for acquiring a patient's multi-lead electrocardiogram; a signal curve fluctuation analysis module for obtaining the degree of fluctuation difference between the signal curves corresponding to each lead and other leads by comparing the difference between the local fluctuations in the data in the voltage sequence of each lead and other leads; a signal curve abnormality analysis module for obtaining the normality factor of the signal curve segments corresponding to subsequent multiple cycles by comparing the time range of each signal curve segment in each lead and the correlation between each signal curve segment and all previous signal curve segments; and an early warning module for issuing early warnings for subsequent multi-lead electrocardiograms based on the degree of fluctuation difference and the normality factor. The present invention reduces the interference of muscle tremors and improves the accuracy of automatic early warnings for multi-lead electrocardiograms.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities. Background Art

[0002] A multi-lead ECG is a clinical examination method that records the heart's electrical activity. It provides more detailed and comprehensive ECG information to help patients assess their heart health. It is primarily used to analyze cardiac abnormalities, determine a patient's response to medication, and reveal trends and changes in cardiac function. Therefore, analyzing multi-lead ECGs is crucial for identifying cardiac abnormalities.

[0003] During the process of acquiring a multi-lead electrocardiogram, the acquired multi-lead electrocardiogram is interfered with by noise due to the interference of muscle tremors. Therefore, when the multi-lead electrocardiogram is used to automatically warn the heart, the accuracy of the automatic warning will be reduced. Summary of the Invention

[0004] The present invention provides a multi-lead electrocardiogram abnormal full-automatic auxiliary analysis system to solve the existing problems.

[0005] The present invention provides a multi-lead electrocardiogram abnormality fully automatic auxiliary analysis system using the following technical solutions:

[0006] One embodiment of the present invention provides a fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities, which includes the following modules:

[0007] A data acquisition module for acquiring a multi-lead electrocardiogram of a patient;

[0008] The signal curve fluctuation analysis module is used to obtain the voltage sequence of each lead through a multi-lead electrocardiogram, and obtain the degree of fluctuation difference between each lead and the corresponding signal curves of other leads by comparing the fluctuation of each data in the local range between each lead and the voltage sequence of other leads;

[0009] The signal curve anomaly analysis module is used to select the R wave peak point from the extreme value points in the signal curve corresponding to each lead, and divide the signal curve corresponding to each lead into several signal curve segments based on the R wave peak point; obtain the normal factor of each signal curve segment in each lead based on the time range of the signal curve segment and the correlation between each signal curve segment and all previous signal curve segments; and predict the normal factors of the signal curve segments corresponding to the subsequent multiple cycles based on the normal factors of all signal curve segments in each lead;

[0010] The early warning module is used to issue early warnings for subsequent multi-lead electrocardiograms based on the degree of fluctuation difference between the corresponding signal curves of each lead and other leads, and the normal factors of the corresponding signal curve segments of subsequent multiple cycles.

[0011] Furthermore, the step of obtaining a voltage sequence of each lead through a multi-lead electrocardiogram includes:

[0012] by The time interval is 1 second. The signal curve of each lead in the electrocardiogram is continuously sampled to obtain the voltage at all times. The voltage is then organized into a set of sequences in chronological order, which is recorded as the voltage sequence of each lead.

[0013] in, The first parameter is preset.

[0014] Furthermore, obtaining the degree of fluctuation difference between each lead and the corresponding signal curves of other leads by comparing the fluctuation of each data in the voltage sequence of each lead with that of other leads within a local range includes:

[0015] Taking the voltage data at each moment in the voltage sequence of each lead as the center, determine the local range at each moment according to a preset range, and record the data corresponding to the local range as the local voltage data at each moment;

[0016] The mean value of the difference between the voltage data at each moment in the voltage sequence of each lead and all local voltage data in the corresponding local range is recorded as the local fluctuation factor of the voltage data at each moment in the voltage sequence of each lead;

[0017] A set of sequences consisting of the local fluctuation factors of the voltage data of each lead at all times is recorded as the local fluctuation sequence of each lead;

[0018] The mean of the differences between all data at the same position in the local fluctuation sequence of each lead and any other lead is recorded as the first difference factor between each lead and any other lead, and the mean of the first difference factors between each lead and all other leads is recorded as the degree of fluctuation difference between the corresponding signal curves of each lead and other leads.

[0019] Furthermore, selecting the R wave peak point from the extreme points in the signal curve corresponding to each lead includes:

[0020] Obtain all the maximum points in the signal curve corresponding to each lead in the electrocardiogram. Starting from the first maximum point, select the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points and record it as the R wave peak point. Then, starting from the fourth maximum point, select the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points and record it as the R wave peak point. And so on to obtain all the R wave peak points.

[0021] Furthermore, the normal factor of each signal curve segment in each lead is obtained by using the time range of the signal curve segment and the correlation between each signal curve segment and all previous signal curve segments, including:

[0022] According to the difference between the time interval between the two ends of each signal curve segment in each lead and the normal time interval, the abnormality degree of each signal curve segment is obtained;

[0023] According to the correlation between each signal curve segment and all previous signal curve segments, the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead is obtained;

[0024] Obtaining a normal factor for each signal curve segment in each lead by determining the correlation between the abnormality degree and the trend distribution;

[0025] Among them, the abnormality degree is negatively correlated with the normal factor, and the trend distribution correlation degree is positively correlated with the normal factor.

[0026] Furthermore, obtaining the abnormality degree of each signal curve segment according to the difference between the time interval between the two ends of each signal curve segment in each lead and the normal time interval includes:

[0027] pass Calculate the heartbeat corresponding to each signal curve segment, record the time interval corresponding to the heartbeat being within the normal heartbeat range as the normal time interval, and record the time interval corresponding to the heartbeat not being within the normal heartbeat range as the abnormal time interval;

[0028] The normal time interval range is obtained through the normal heartbeat range, and the difference between each abnormal time interval and the median value within the normal time interval range is recorded as the abnormality degree of the signal curve segment corresponding to each abnormal time interval; wherein the abnormality degree of the signal curve segment corresponding to each normal time interval is recorded as 0;

[0029] in, Indicates the time interval corresponding to each signal curve segment.

[0030] Furthermore, obtaining the trend distribution correlation degree between each signal curve segment in each lead and all previous signal curve segments based on the correlation between each signal curve segment and all previous signal curve segments includes:

[0031] The result of normalizing the correlation coefficient between each signal curve segment in each lead and any previous signal curve segment is recorded as the first correlation coefficient between each signal curve segment and any previous signal curve segment;

[0032] The result of negative correlation mapping and normalization of the difference in the number of signal curve segments between each signal curve segment in each lead and any previous signal curve segment is recorded as the weight between each signal curve segment and any previous signal curve segment;

[0033] The first correlation coefficients between each signal curve segment and all previous signal curve segments are weighted and averaged by the weights to obtain the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead.

[0034] Furthermore, predicting the normal factors of the signal curve segments corresponding to the subsequent multiple cycles by using the normal factors of all signal curve segments in each lead includes:

[0035] The normal factors of all signal curve segments in each lead are sorted in chronological order to form a set of sequences, which are recorded as the normal factor sequence of each lead. The normal factors of the corresponding signal curve segments in the subsequent multiple cycles are predicted using the ARIMA model.

[0036] Furthermore, the method of providing an early warning for subsequent multi-lead electrocardiograms by measuring the degree of fluctuation difference between the signal curves corresponding to each lead and other leads and the normal factors of the signal curve segments corresponding to subsequent cycles includes:

[0037] According to the fluctuation difference between each lead and the corresponding signal curves of other leads and the normal factor of the signal curve segment corresponding to each subsequent cycle of each lead, the first abnormality degree of the signal curve segment corresponding to each subsequent cycle of each lead is obtained; the average of the first abnormality degrees of the signal curve segment corresponding to each subsequent cycle of all leads is recorded as the abnormality factor of the signal curve segment corresponding to each subsequent cycle;

[0038] The abnormal factor of the signal curve segment corresponding to each subsequent cycle is greater than the preset threshold When the abnormal factor of the corresponding signal curve segment in each subsequent cycle is less than or equal to the preset threshold If the alarm is set to 0, it is a normal situation and no warning is needed.

[0039] Furthermore, obtaining a first abnormality degree of a signal curve segment corresponding to each subsequent cycle of each lead according to a fluctuation difference degree between the signal curves corresponding to each lead and other leads and a normal factor of a signal curve segment corresponding to each subsequent cycle of each lead includes:

[0040] Perform negative correlation normalization mapping on the fluctuation difference between each lead and the corresponding signal curves of other leads, and record the result as the first coefficient of each lead; normalize the normal factor of the signal curve segment corresponding to each subsequent cycle of each lead to the inverse, and record the result as the second abnormality degree of the signal curve segment corresponding to each subsequent cycle of each lead;

[0041] The second abnormality level is adjusted using the first coefficient to obtain a first abnormality level of a signal curve segment corresponding to each subsequent cycle of each lead.

[0042] The beneficial effects of the technical solution of the present invention are as follows: the present invention obtains the degree of fluctuation difference between the signal curves corresponding to each lead and other leads by the difference between the fluctuations of each data in the voltage sequence of each lead and other leads in a local range, thereby improving the accuracy of muscle interference analysis on the signal curves of individual leads in a multi-lead electrocardiogram; obtains the normal factor of each signal curve segment in each lead by the time range of each signal curve segment in each lead and the correlation between each signal curve segment and all previous signal curve segments, thereby improving the accuracy of analysis of periodic changes of signal curves in a multi-lead electrocardiogram; predicts the normal factors of signal curve segments corresponding to subsequent multiple cycles by the normal factors of all signal curve segments in each lead; automatically warns of subsequent multi-lead electrocardiograms by the degree of fluctuation difference between the signal curves corresponding to each lead and other leads and the normal factors of signal curve segments corresponding to subsequent multiple cycles, that is, the influence of muscle tremor on the interference of normal factors is reduced by the said degree of fluctuation difference, thereby improving the accuracy of automatic warning of multi-lead electrocardiograms. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 This is a module flow chart of a fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to the present invention;

[0045] Figure 2 Schematic diagram of the composition of an electrocardiogram. DETAILED DESCRIPTION

[0046] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities proposed in accordance with the present invention. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0047] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0048] The specific scheme of the multi-lead electrocardiogram abnormality fully automatic auxiliary analysis system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0049] See also Figure 1 , which shows a module flow chart of a multi-lead electrocardiogram abnormality fully automatic auxiliary analysis system provided by one embodiment of the present invention, the system includes the following modules:

[0050] Module 101: a data acquisition module, used to obtain a multi-lead electrocardiogram of a patient.

[0051] It should be noted that in order to analyze the real-time condition of the heart and provide early warning for the heart, multi-lead electrocardiograms are collected for analysis.

[0052] Specifically, a multi-lead electrocardiogram of the patient is collected within one week before the current moment, thereby obtaining the patient's multi-lead electrocardiogram; wherein, in this embodiment, the multi-lead electrocardiogram is analyzed using twelve leads as an example. The duration of the multi-lead electrocardiogram collection is not specifically limited in this embodiment and can be determined by the implementer according to specific circumstances.

[0053] At this point, the patient's multi-lead electrocardiogram is obtained.

[0054] Module 102: A signal curve fluctuation analysis module, which is used to obtain the voltage sequence of each lead through a multi-lead electrocardiogram, and obtain the degree of fluctuation difference between each lead and the corresponding signal curves of other leads by comparing the fluctuation of each data in the voltage sequence of each lead with that of other leads within a local range.

[0055] It should be noted that when analyzing cardiac early warning through electrocardiogram, the abnormalities of the heart include: myocardial ischemia, myocardial infarction, myocardial damage, atrial and ventricular fibrillation, hypercardia and bradycardia, and atrioventricular and ventricular premature beats, etc. Therefore, the above situations are roughly divided into several categories for analysis. Hypercardia and bradycardia are the faster and slower heartbeats. Myocardial ischemia, myocardial infarction, and myocardial damage correspond to the positive and negative of different waves in the electrocardiogram. Atrioventricular and ventricular premature beats may destroy the integrity of the periodic QRS wave, etc. Therefore, the analysis is carried out through roughly several categories of situations. There are roughly several types of waves in the electrocardiogram, namely P wave, QRS wave and T wave. A P wave, QRS wave and T wave can be used to determine a heartbeat cycle, that is, a complete process of the beating of myocardial cells in the heart. Under normal circumstances, there is a P wave first, followed by a QRS wave, and finally a T wave. The composition of the electrocardiogram in one cycle is as follows. Figure 2 shown.

[0056] It should be further explained that since a multi-lead ECG is obtained by measuring voltage changes at different locations within the heart's myocardial cells, when a heart abnormality occurs, all signal curves in the multi-lead ECG will change, but the direction of the changes will differ. That is, the voltage of the myocardial cells changes at the same time in all signal curves in the multi-lead ECG, resulting in some curves fluctuating upward and others downward. Therefore, under normal circumstances, all signal curves in a multi-lead ECG will either fluctuate or not at all at the same time. Therefore, the degree of cardiac abnormality can be further analyzed by analyzing the fluctuations of different signal curves at the same time. However, in a multi-lead ECG, multiple leads are composed of limb leads and chest leads. Therefore, when muscles in the body undergo localized fibrillation, this typically manifests as changes in individual signal curves rather than changes across all signal curves. Therefore, the degree of fluctuation difference between each lead and its corresponding signal curves is determined by comparing the fluctuations of each data point within the voltage series of each lead with those of the other leads.

[0057] Preferably, a first parameter is preset , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. It can be determined according to the specific implementation situation. The time interval is seconds. The signal curve corresponding to each lead in the electrocardiogram is continuously sampled to obtain the voltage at all times. A set of sequences is formed in chronological order, which is recorded as the voltage sequence of each lead.

[0058] Furthermore, as an embodiment, the specific calculation method of the fluctuation difference between each lead and the corresponding signal curves of other leads is as follows:

[0059] Centered on the voltage data at each moment in the voltage sequence of each lead, according to the preset range Determine a local range at each moment, and record data corresponding to the local range as local voltage data at each moment;

[0060] The mean value of the difference between the voltage data at each moment in the voltage sequence of each lead and all local voltage data in the corresponding local range is recorded as the local fluctuation factor of the voltage data at each moment in the voltage sequence of each lead;

[0061] A set of sequences consisting of the local fluctuation factors of the voltage data of each lead at all times is recorded as the local fluctuation sequence of each lead;

[0062] The mean of the differences between all data at the same position in the local fluctuation sequence of each lead and any other lead is recorded as the first difference factor between each lead and any other lead, and the mean of the first difference factors between each lead and all other leads is recorded as the degree of fluctuation difference between the corresponding signal curves of each lead and other leads.

[0063] In one embodiment of the present invention, it is specifically expressed by the formula:

[0064]

[0065]

[0066] Where, Indicates the voltage sequence of each lead. Voltage data at a moment, Indicates the voltage sequence of each lead. The local range of the moment Local voltage data, is the absolute value symbol, Indicates the number of all local voltage data within the local range at each moment, Indicates the voltage sequence of each lead. The local fluctuation factor of the voltage data at a certain moment, Indicates that each lead is The first local fluctuation sequence of the lead The difference between the data, Represents the number of all data in the local fluctuation sequence, Indicates the number of all leads, Indicates that each lead is The degree of fluctuation difference between the signal curves corresponding to the leads. The local fluctuation sequence of each lead is a set of sequences composed of the local fluctuation factors of the voltage data of each lead at all moments in time.

[0067] Among them, the greater the difference between the data in the local fluctuation sequence of each lead and all other leads, the more different the data fluctuation of the signal curve corresponding to the lead and all other leads at the same time is, that is, the greater the possibility that the signal curve of the lead is affected by muscle tremor and causes abnormal fluctuation; conversely, the more similar the data fluctuation of the signal curve corresponding to the lead and all other leads at the same time is, that is, the less likely the signal curve of the lead is affected by muscle tremor and causes abnormal fluctuation.

[0068] in, To preset the second parameter, in this embodiment This example is described as an example, and this embodiment is not specifically limited. It may depend on the specific implementation situation.

[0069] At this point, the degree of fluctuation difference between each lead and the corresponding signal curves of other leads is obtained.

[0070] Module 103: A signal curve anomaly analysis module, configured to select R-wave peak points from the extreme points in the signal curve corresponding to each lead, and divide the signal curve corresponding to each lead into a number of signal curve segments based on the R-wave peak points; obtain the normal factor of each signal curve segment in each lead based on the time range of the signal curve segment and the correlation between each signal curve segment and all previous signal curve segments; and predict the normal factors of the signal curve segments corresponding to the subsequent multiple cycles based on the normal factors of all signal curve segments in each lead.

[0071] It should be noted that since the fluctuations of each signal curve in the electrocardiogram are cyclical, that is, the P wave, QRS wave, and T wave change periodically in this order, when there is no abnormality in the heart, the fluctuations of each signal curve will continue to show cyclical changes. Therefore, the signal curve segments corresponding to all cycles can be obtained through the peaks, and the differences between the signal curve segments can be used to analyze whether there are abnormalities in the electrocardiogram. Therefore, the R wave peak point can be selected from the extreme points in the signal curve corresponding to each lead, and the signal curve corresponding to each lead can be divided into several signal curve segments based on the R wave peak point.

[0072] Preferably, all the maximum points in the signal curve corresponding to each lead in the electrocardiogram are obtained. Starting from the first maximum point, the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points is selected and recorded as the R wave peak point. Then, starting from the fourth maximum point, the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points is selected and recorded as the R wave peak point. In this way, all R wave peak points are obtained. If there are fewer than three maximum points at the end, no R wave peak point selection is performed. The signal curve corresponding to each lead is divided into a plurality of signal curve segments by the R wave peak points. Thus, a plurality of signal curve segments on the signal curve corresponding to each lead are obtained.

[0073] It should be further explained that when the heart is normal, the heartbeat rate does not vary much per minute. In other words, the intervals between all adjacent R-wave peaks in the electrocardiogram are the same. However, when the heart is abnormal, the heartbeat rate can vary significantly. Therefore, the abnormality can be analyzed by analyzing the changes in the heartbeat rate between adjacent cycles. Furthermore, because the distribution of all adjacent signal curve segments in each lead's signal curve is similar under normal circumstances, the normality factor of each signal curve segment in each lead is obtained by analyzing the time range of each signal curve segment in each lead and the correlation between each signal curve segment and all previous signal curve segments.

[0074] Preferably, as an embodiment, the specific calculation method of the normal factor of each signal curve segment in each lead is:

[0075] According to the difference between the time interval between the two ends of each signal curve segment in each lead and the normal time interval, the abnormality degree of each signal curve segment is obtained;

[0076] According to the correlation between each signal curve segment and all previous signal curve segments, the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead is obtained;

[0077] Obtaining a normal factor for each signal curve segment in each lead by determining the correlation between the abnormality degree and the trend distribution;

[0078] Among them, the abnormality degree is negatively correlated with the normal factor, and the trend distribution correlation degree is positively correlated with the normal factor.

[0079] In one embodiment of the present invention, it is specifically expressed by the formula:

[0080]

[0081] Where, Indicates the first The degree of correlation between the trend distribution of a signal curve segment and all previous signal curve segments, represents an exponential function with a natural constant as the base, Indicates the The abnormality of each signal curve segment, Indicates the first Normal factor for each signal curve segment.

[0082] Among them, when the trend distribution correlation is greater and the abnormality is smaller, it means that the normal factor of the signal curve segment is larger; otherwise, it means that the normal factor of the signal curve segment is smaller.

[0083] The process of obtaining the abnormality degree of each signal curve segment in each lead is as follows:

[0084] Preferably, the time interval between the two ends of each signal curve segment on the signal curve corresponding to each lead is obtained; Calculate whether the heartbeat corresponding to each signal curve segment is within the normal heartbeat range. If the heartbeat is within the normal range, the time interval corresponding to the signal curve segment is recorded as the normal time interval. If the heartbeat is not within the normal range, the time interval corresponding to the signal curve segment is recorded as the abnormal time interval. Represents the time interval corresponding to each signal curve segment. This is a formula for calculating heartbeat, which is a conventionally known formula.

[0085] The normal heartbeat range is used to obtain the normal time interval range. The difference between each abnormal time interval and the median value within the normal time interval range is recorded as the abnormality level of the signal curve segment corresponding to each abnormal time interval. The abnormality level of the signal curve segment corresponding to each normal time interval is recorded as 0. Thus, the abnormality level of each signal curve segment in each lead is obtained. The obtained normal time interval range is the normal time interval range corresponding to each signal curve segment.

[0086] The process of obtaining the trend distribution correlation degree between each signal curve segment in each lead and all previous signal curve segments is as follows:

[0087] Preferably, as an embodiment, the specific calculation method of the trend distribution correlation degree between each signal curve segment in each lead and all previous signal curve segments is:

[0088] The result of normalizing the correlation coefficient between each signal curve segment in each lead and any previous signal curve segment is recorded as the first correlation coefficient between each signal curve segment and any previous signal curve segment;

[0089] The result of negative correlation mapping and normalization of the difference in the number of signal curve segments between each signal curve segment in each lead and any previous signal curve segment is recorded as the weight between each signal curve segment and any previous signal curve segment;

[0090] The first correlation coefficients between each signal curve segment and all previous signal curve segments are weighted and averaged by the weights to obtain the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead.

[0091] In one embodiment of the present invention, it is specifically expressed by the formula:

[0092]

[0093] Where, Indicates the first The signal curve segment is consistent with the previous The Spearman correlation coefficient between the signal curve segments, For each lead The signal curve segment is consistent with the previous The difference between the signal curve segments is the reciprocal of the number of signal curve segments plus 1, which represents the weight between the two signal curve segments. Indicates the first The signal curve segment is consistent with the previous The phase difference between the signal curve segments is the reciprocal of the number of signal curve segments plus 1. Indicates the number of all signal curve segments before each signal curve segment, Indicates the first The degree of correlation between the trend distribution of a signal curve segment and all previous signal curve segments, Represents a linear normalization function. The Spearman correlation coefficient is a well-known technique and will not be described in detail here.

[0094] Among them, the greater the correlation between each signal curve segment and all previous signal curve segments, the more normal the fluctuation and distribution of each lead are, and the more consistent with the periodic change; conversely, the less consistent with the periodic change.

[0095] It should be noted that, since abnormal heartbeats and peaks will show trend changes, the normal factors of all signal curve segments in each lead are used to predict the normal factors of the corresponding signal curve segments in subsequent cycles.

[0096] Preferably, the normal factors of all signal curve segments in each lead are sorted in chronological order to form a sequence, recorded as the normal factor sequence for each lead. The ARIMA model is then used to predict the normal factors of the corresponding signal curve segments for multiple subsequent cycles. The ARIMA model is well known and will not be described in detail here.

[0097] Module 104: an early warning module, configured to provide early warning for subsequent multi-lead electrocardiograms based on the degree of fluctuation difference between the signal curves corresponding to each lead and other leads and the normal factors of the signal curve segments corresponding to subsequent multiple cycles.

[0098] It should be noted that the greater the fluctuation difference between each lead and the corresponding signal curve of the other leads, the greater the possibility that the signal curve of the lead is abnormal due to muscle tremor, that is, the lower the reliability of the detected abnormality; the smaller the fluctuation difference between each lead and the corresponding signal curve of the other leads, the less likely the signal curve of the lead is abnormal due to muscle tremor, that is, the higher the reliability of the detected abnormality; the smaller the impact of muscle tremor, the more reliable the abnormality or normality of the analyzed lead. Therefore, the fluctuation difference between each lead and the corresponding signal curve of the other leads and the normal factors of the corresponding signal curve segments of subsequent cycles can be used to provide early warning for subsequent multi-lead electrocardiograms.

[0099] Preferably, as an embodiment, the specific calculation method of the abnormality factor of the signal curve segment corresponding to each subsequent period is:

[0100] Perform negative correlation normalization mapping on the fluctuation difference between each lead and the corresponding signal curves of other leads, and record the result as the first coefficient of each lead; normalize the normal factor of the signal curve segment corresponding to each subsequent cycle of each lead to the inverse, and record the result as the second abnormality degree of the signal curve segment corresponding to each subsequent cycle of each lead;

[0101] The second abnormality degree is adjusted by the first coefficient to obtain a first abnormality degree of a signal curve segment corresponding to each subsequent cycle of each lead;

[0102] The average of the first abnormality degree of the signal curve segment corresponding to each subsequent cycle of all leads is recorded as the abnormality factor of the signal curve segment corresponding to each subsequent cycle.

[0103] In one embodiment of the present invention, it is specifically expressed by the formula:

[0104]

[0105] Where, Indicates the The degree of fluctuation difference between the signal curves of a lead and other leads, Indicates the The subsequent The normal factor of the signal curve segment corresponding to each cycle, Indicates the number of all leads, Indicates the subsequent The abnormal factor of the signal curve segment corresponding to each cycle, represents an exponential function with a natural constant as the base, represents the linear normalization function.

[0106] Among them, the greater the degree of fluctuation difference, the lower the reliability of the abnormal and normal conditions in the signal curve, and vice versa, the higher the reliability; when each cycle of all subsequent leads corresponds to the normal factor of the signal curve segment, the cycle corresponds to the abnormal factor of the signal curve segment.

[0107] Preset a threshold , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. It may depend on the specific implementation situation.

[0108] When the abnormal factor of the signal curve segment corresponding to each subsequent cycle is greater than the preset threshold When the abnormal factor of the corresponding signal curve segment in each subsequent cycle is less than or equal to the preset threshold No warning is required.

[0109] It should be noted that the The model is only used to represent negative correlation and constrain the output of the model to be in In the specific implementation, it can be replaced by other models with the same purpose. This embodiment is only based on The model is described as an example without any specific limitation. is the input to the model.

[0110] At this point, this embodiment is completed.

[0111] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities, characterized in that: The system includes the following modules: A data acquisition module for acquiring a multi-lead electrocardiogram of a patient; A signal curve fluctuation analysis module is used to obtain the voltage sequence of each lead through a multi-lead electrocardiogram, and obtain the degree of fluctuation difference between the signal curves corresponding to each lead and other leads by comparing the fluctuation of each data in the local range between each lead and the voltage sequences of other leads, including: taking the voltage data at each moment in the voltage sequence of each lead as the center, determining the local range of each moment according to a preset range, and recording the data corresponding to the local range as the local voltage data at each moment; recording the average value of the difference between the voltage data at each moment in the voltage sequence of each lead and all local voltage data in the corresponding local range as the local fluctuation factor of the voltage data at each moment in the voltage sequence of each lead; recording a group of sequences consisting of the local fluctuation factors of the voltage data at all moments of each lead as the local fluctuation sequence of each lead; recording the average value of the difference between all data at the same position in the local fluctuation sequence of each lead and any other lead as the first difference factor between each lead and any other lead, and recording the average value of the first difference factors between each lead and all other leads as the degree of fluctuation difference between the signal curves corresponding to each lead and other leads; The signal curve anomaly analysis module is used to select the R wave peak point from the extreme value points in the signal curve corresponding to each lead, and divide the signal curve corresponding to each lead into several signal curve segments based on the R wave peak point; obtain the normal factor of each signal curve segment in each lead based on the time range of the signal curve segment and the correlation between each signal curve segment and all previous signal curve segments; and predict the normal factors of the signal curve segments corresponding to the subsequent multiple cycles based on the normal factors of all signal curve segments in each lead; The early warning module is used to issue early warnings for subsequent multi-lead electrocardiograms based on the degree of fluctuation difference between the corresponding signal curves of each lead and other leads, and the normal factors of the corresponding signal curve segments of subsequent multiple cycles.

2. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 1, characterized in that: The method of obtaining a voltage sequence of each lead through a multi-lead electrocardiogram includes: by The time interval is 1 second. The signal curve of each lead in the electrocardiogram is continuously sampled to obtain the voltage at all times. The voltage is then organized into a set of sequences in chronological order, which is recorded as the voltage sequence of each lead. in, The first parameter is preset.

3. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 1, characterized in that: The step of selecting the R wave peak point from the extreme points in the signal curve corresponding to each lead includes: Obtain all the maximum points in the signal curve corresponding to each lead in the electrocardiogram. Starting from the first maximum point, select the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points and record it as the R wave peak point. Then, starting from the fourth maximum point, select the maximum point with the largest value corresponding to the vertical axis among the three adjacent maximum points and record it as the R wave peak point. And so on to obtain all the R wave peak points.

4. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 1, characterized in that: The method of obtaining a normal factor of each signal curve segment in each lead by using a time range of the signal curve segment and a correlation between each signal curve segment and all previous signal curve segments includes: According to the difference between the time interval between the two ends of each signal curve segment in each lead and the normal time interval, the abnormality degree of each signal curve segment is obtained; According to the correlation between each signal curve segment and all previous signal curve segments, the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead is obtained; Obtaining a normal factor for each signal curve segment in each lead by determining the correlation between the abnormality degree and the trend distribution; Among them, the abnormality degree is negatively correlated with the normal factor, and the trend distribution correlation degree is positively correlated with the normal factor.

5. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 4, characterized in that: The step of obtaining the abnormality degree of each signal curve segment according to the difference between the time interval between the two ends of each signal curve segment in each lead and the normal time interval includes: pass Calculate the heartbeat corresponding to each signal curve segment, record the time interval corresponding to the heartbeat being within the normal heartbeat range as the normal time interval, and record the time interval corresponding to the heartbeat not being within the normal heartbeat range as the abnormal time interval; The normal time interval range is obtained through the normal heartbeat range, and the difference between each abnormal time interval and the median value within the normal time interval range is recorded as the abnormality degree of the signal curve segment corresponding to each abnormal time interval; wherein the abnormality degree of the signal curve segment corresponding to each normal time interval is recorded as 0; in, Indicates the time interval corresponding to each signal curve segment.

6. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 4, characterized in that: The step of obtaining the trend distribution correlation degree between each signal curve segment in each lead and all previous signal curve segments according to the correlation between each signal curve segment and all previous signal curve segments includes: The result of normalizing the correlation coefficient between each signal curve segment in each lead and any previous signal curve segment is recorded as the first correlation coefficient between each signal curve segment and any previous signal curve segment; The result of negative correlation mapping and normalization of the difference in the number of signal curve segments between each signal curve segment in each lead and any previous signal curve segment is recorded as the weight between each signal curve segment and any previous signal curve segment; The first correlation coefficients between each signal curve segment and all previous signal curve segments are weighted and averaged by the weights to obtain the trend distribution correlation degree between each signal curve segment and all previous signal curve segments in each lead.

7. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 1, characterized in that: The method of predicting the normal factors of the signal curve segments corresponding to the subsequent multiple cycles by using the normal factors of all signal curve segments in each lead includes: The normal factors of all signal curve segments in each lead are sorted in chronological order to form a set of sequences, which are recorded as the normal factor sequence of each lead. The normal factors of the corresponding signal curve segments in the subsequent multiple cycles are predicted using the ARIMA model.

8. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 1, characterized in that: The method provides an early warning for subsequent multi-lead electrocardiograms by determining the degree of fluctuation difference between the signal curves corresponding to each lead and other leads and the normal factors of the signal curve segments corresponding to subsequent cycles, including: According to the fluctuation difference between each lead and the corresponding signal curves of other leads and the normal factor of the signal curve segment corresponding to each subsequent cycle of each lead, the first abnormality degree of the signal curve segment corresponding to each subsequent cycle of each lead is obtained; the average of the first abnormality degrees of the signal curve segment corresponding to each subsequent cycle of all leads is recorded as the abnormality factor of the signal curve segment corresponding to each subsequent cycle; The abnormal factor of the signal curve segment corresponding to each subsequent cycle is greater than the preset threshold When the abnormal factor of the corresponding signal curve segment in each subsequent cycle is less than or equal to the preset threshold If the alarm is set to 0, it is a normal situation and no warning is needed.

9. The fully automatic auxiliary analysis system for multi-lead electrocardiogram abnormalities according to claim 8, characterized in that: The method of obtaining a first abnormality degree of a signal curve segment corresponding to each subsequent cycle of each lead according to a fluctuation difference degree between the signal curves corresponding to each lead and other leads and a normal factor of a signal curve segment corresponding to each subsequent cycle of each lead includes: Perform negative correlation normalization mapping on the fluctuation difference between each lead and the corresponding signal curves of other leads, and record the result as the first coefficient of each lead; normalize the inverse of the normal factor of the signal curve segment corresponding to each subsequent cycle of each lead, and record the result as the second abnormality degree of the signal curve segment corresponding to each subsequent cycle of each lead; The second abnormality level is adjusted using the first coefficient to obtain a first abnormality level of a signal curve segment corresponding to each subsequent cycle of each lead.

Images