An electrocardiogram anti-interference detection method and system based on big data analysis

Abstract

The application relates to an electrocardiogram anti-interference detection method and system based on big data analysis, which comprises the following steps: an ECG signal output by an electrocardiograph is received in real time and input into a pre-trained interference adjustment model; the interference adjustment model extracts the ECG signal and generates signal segments of different time sequences based on a preset sampling rule; the interference adjustment model performs feature recognition on the multiple signal segments, judges whether there is a signal segment disturbed by interference, and marks the signal segment as an abnormal segment; when the abnormal segment exists, the abnormal segment and other signal segments associated with the abnormal segment in time sequence are acquired to form a feature set; the interference type of the feature set is recognized and judged, a corresponding adjustment strategy is screened out based on the recognized interference type, error adjustment is performed on the signal segments needing adjustment in the feature set based on the adjustment strategy, and an electrocardiogram is generated and sent to a user end. The application has the effect of improving the accuracy of electrocardiogram interference signal troubleshooting.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electrocardiogram detection, in particular to an electrocardiogram anti-interference detection method and system based on big data analysis. BACKGROUND

[0002] At present, in the ECG signal recognition process of electrocardiogram, the electrocardio signal is easily disturbed by various noises, which may cover the real physiological signal and affect the diagnostic accuracy.

[0003] Common types of electrocardio interference include power frequency interference, baseline drift, electromyographic interference and partial high-frequency noise interference. For various types of electrocardio interference, the current common method is to use a filter corresponding to the frequency for filtering processing. However, the filtering parameters between different filters are relatively fixed, which is prone to signal distortion and poor adaptability to dynamic noise, resulting in low accuracy of the detection result, which needs to rely on human experience for judgment, and therefore the anti-interference detection method needs to be improved. SUMMARY

[0004] In order to improve the accuracy of electrocardiogram interference signal investigation and improve the accuracy of electrocardiogram detection results, the present application provides an electrocardiogram anti-interference detection method and system based on big data analysis.

[0005] The above invention of the present application is achieved by the following technical scheme:

[0006] An electrocardiogram anti-interference detection method based on big data analysis, comprising the steps of:

[0007] receiving the ECG signal output by the electrocardio monitor in real time and inputting it into the interference adjustment model trained in advance;

[0008] The interference adjustment model extracts the ECG signal and generates signal segments of different time sequences based on the preset sampling rule;

[0009] The interference adjustment model performs feature recognition on the plurality of signal segments, judges whether there is a signal segment disturbed by interference, and marks it as an abnormal segment;

[0010] When there is an abnormal segment, the abnormal segment and other signal segments associated with the abnormal segment in time sequence are obtained to form a feature set;

[0011] The interference type of the feature set is recognized and judged, and the corresponding adjustment strategy is selected based on the recognized interference type, the error of the signal segment needing adjustment in the feature set is adjusted based on the adjustment strategy, and the electrocardiogram is generated and sent to the user end.

[0012] By adopting the technical scheme, the ECG signal detected in real time is received through the interference adjustment model first, then a plurality of ECG signal segments, i.e., signal segments, are obtained based on a preset sampling rule, the plurality of continuous signal segments are identified to determine whether the ECG signal has an abnormal waveform, an abnormal segment is obtained, when the abnormal segment exists, the specific interference type of the abnormal segment needs to be further confirmed, and then the waveforms in the ECG signal can be filtered or / and calibrated more accurately, therefore, when the interference type is confirmed, the other signal segments associated with the time sequence of the abnormal segment need to be combined, i.e., the waveform features before and after the abnormal segment are combined, the interference type is determined through the change of the waveform trend in the signal segment, the power frequency interference, baseline drift, electromyographic interference and partial high-frequency noise interference, finally, the corresponding adjustment strategy is called to generate an electrocardiogram by adjusting the error of the signal segment and sending the electrocardiogram to the user end; therefore, through the machine learning capability of the interference adjustment model, the abnormal ECG signal is accurately determined, identified and adjusted by combining the waveform features before and after the ECG signal, and the detection result is more accurate.

[0013] Optionally, the step of extracting the ECG signal and generating signal segments with different time sequences based on the preset sampling rule by the interference adjustment model comprises:

[0014] The interference adjustment model labels the received ECG signal based on a preset interval duration to obtain a plurality of marking points;

[0015] The ECG signal between adjacent n marking points is sampled and extracted to generate a signal segment, and n is a positive integer;

[0016] The time information of the n marking points in the signal segment is obtained as the time sequence information associated with the signal segment.

[0017] By adopting the technical scheme, the interference adjustment model automatically labels the ECG signal received in real time based on the preset interval duration to obtain a plurality of marking points, and the ECG signal is sliced by the number of marking points, compared with the sampling mode based on the interval duration, the acquisition of the signal segment is more flexible and convenient, and the matching between the time sequence information and the signal segment is facilitated by the marking points.

[0018] Optionally, the step of identifying the features of a plurality of signal segments, determining whether there is a signal segment disturbed by interference, and marking the signal segment as an abnormal segment by the interference adjustment model comprises:

[0019] The interference adjustment model identifies the features of a preset number of signal segments at a time according to the generation order of the signal segments;

[0020] Based on image feature comparison, the deviation wave band existing in the signal segments is identified and marked, and the waveform change coefficient of the deviation wave band is calculated to determine whether there is an abnormal segment in the current signal segments.

[0021] By adopting the above technical solution, in order to more accurately determine whether the signal segment is abnormal, the interference adjustment model adopts single-time identification on multiple continuous signal segments of time sequence information, judges and marks the deviation wave band through image comparison of continuous waveform changes, calculates the waveform change coefficient of the wave band, that is, the degree of deviation, and determines whether it belongs to the normal fluctuation range or exists interference through the degree of deviation.

[0022] Optionally, in the step of identifying and marking the deviation wave band existing in the signal segments based on image feature comparison, calculating the waveform change coefficient of the deviation wave band, and determining whether there is an abnormal segment in the current signal segments, the step includes:

[0023] Obtaining an overall waveform image formed by the signal segments;

[0024] Through the preset plane coordinate system, the overall waveform image is compared with the pre-stored reference image in image features, and the deviation wave band existing in the reference image is marked;

[0025] Obtaining the coordinate points of the deviation wave band, and calculating the waveform change coefficient of the deviation wave band;

[0026] Identifying the position of the deviation wave band in the overall waveform image to determine the wave band type and signal segment to which the deviation wave band belongs;

[0027] Matching the wave band type with the preset coefficient threshold value, and comparing the calculated waveform change coefficient, when the waveform change coefficient is greater than or equal to the coefficient threshold value, determining that the signal segment to which the current deviation wave band belongs is an abnormal segment, and when the waveform change coefficient is less than the coefficient threshold value, determining that the current deviation wave band is a normal fluctuation range.

[0028] By adopting the above technical solution, by combining the signal segments to form an overall waveform image and mapping it into a preset coordinate system, the deviation wave band can be marked while the end point and inflection point coordinates in the deviation wave band are obtained, so as to facilitate the calculation of the waveform change coefficient, and after the overall waveform image is formed, the wave band order of the ECG signal has a regularity, so that the recognition of the wave band type of the deviation wave band is more accurate, so that the coefficient threshold value of the corresponding wave band can be found to more accurately determine whether the deviation wave band is a normal fluctuation or is affected by interference. The standard image includes normal electrocardiogram images and pathological electrocardiogram images, such as tachycardia, bradycardia, atrial flutter, atrial and ventricular fibrillation, myocardial ischemia, and myocardial infarction.

[0029] Optionally, in the step of acquiring the abnormal segment and other signal segments associated with the time sequence of the abnormal segment to form a feature set when the abnormal segment exists, the feature set includes the remaining signal segments except the abnormal segment in the interference adjustment model single identification.

[0030] Optionally, the step of identifying and judging the interference type of the feature set, and screening out the corresponding adjustment strategy based on the identified interference type, error adjusting the signal segments that need to be adjusted in the feature set based on the adjustment strategy, and generating an electrocardiogram and sending it to the user end, includes:

[0031] Based on the overall waveform image and the deviation wave band in the feature set, the interference type of the current feature set is matched, and the number of interference types includes one or more;

[0032] Based on the interference type, the corresponding adjustment strategy is screened out to error adjust the deviation wave band in the abnormal segment, and the error adjustment includes filtering and / or calibration processing of the interference signal;

[0033] The signal segment after error adjustment is combined according to the time sequence information to generate an electrocardiogram and send it to the user end.

[0034] By adopting the above technical scheme, through the overall waveform image and the position and waveform of the deviation wave band in the overall waveform image, the interference type matching the current deviation wave band and the position of the deviation wave band can be screened out from the preset database, and the preset adjustment strategy is correspondingly screened out to error adjust the deviation wave band. Error adjustment also includes different adjustment methods, such as baseline calibration, high-frequency interference signal filtering, and waveform adjustment, and finally generates a denoised electrocardiogram to the user end.

[0035] Optionally, the step of matching the interference type of the current feature set based on the overall waveform image and the deviation wave band in the feature set includes:

[0036] Based on the context association rule, the distribution position of the deviation wave band in the overall waveform image is identified; the wave band type and the waveform change coefficient of the deviation wave band are identified one by one;

[0037] The interference type of the current deviation wave band is matched from the preset interference type database one by one as the interference type of the current feature set.

[0038] By adopting the above technical scheme, each interference type in the preset interference type database is mapped with multiple corresponding images through machine learning. When the image features match, that is, the distribution position, the wave band type, and the change features of the waveform itself match, the corresponding interference type can be matched, and there may be multiple deviation wave bands with different features in the same feature set. It is necessary to match one by one to achieve accurate adjustment.

[0039] The second application purpose of the present application is achieved by the following technical solution:

[0040] An electrocardiogram anti-interference detection system based on big data analysis, comprising:

[0041] A signal receiving module is configured to receive an ECG signal output by an electrocardiogram monitor in real time and input the ECG signal into a pre-trained interference adjustment model;

[0042] A sampling module is configured to extract the ECG signal and generate signal segments of different time sequences based on a preset sampling rule;

[0043] An identification module is configured to perform feature identification on the multiple signal segments by the interference adjustment model, determine whether there is a signal segment disturbed by interference, and mark the signal segment as an abnormal segment;

[0044] An association module is configured to, when there is an abnormal segment, acquire the abnormal segment and other signal segments associated with the abnormal segment in terms of time sequence, and form a feature set;

[0045] An adjustment module is configured to identify and determine the interference type of the feature set, filter out a corresponding adjustment strategy based on the identified interference type, perform error adjustment on the signal segments that need to be adjusted in the feature set based on the adjustment strategy, and generate an electrocardiogram and send the electrocardiogram to a user end.

[0046] The third application purpose of the present application is achieved by the following technical solution:

[0047] A computer device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned electrocardiogram anti-interference detection method based on big data analysis when executing the computer program.

[0048] The fourth application purpose of the present application is achieved by the following technical solution:

[0049] A computer readable storage medium stores a computer program, and the computer program implements the steps of the above-mentioned electrocardiogram anti-interference detection method based on big data analysis when executed by a processor.

[0050] In summary, the present application includes at least one of the following beneficial technical effects:

[0051] 1. The method comprises the following steps: identifying a plurality of continuous signal segments, determining whether the ECG signal has abnormal waveform, when there is an abnormal segment, further confirming the specific interference type of the abnormal segment, and then more accurately filtering or / and calibrating the waveform in the ECG signal, so when confirming the interference type, the other signal segments associated with the time sequence of the abnormal segment need to be combined, that is, the waveform characteristics before and after the abnormal segment are combined, the interference type is determined by the change of the waveform trend in the signal segment, and finally the corresponding adjustment strategy is called to generate an electrocardiogram after error adjustment of the signal segment and sent to the user end; therefore, through the machine learning ability of the interference adjustment model, the waveform characteristics before and after the ECG signal are combined to accurately judge, identify and adjust the abnormal ECG signal, so that the detection result is more accurate;

[0052] 2. The interference adjustment model automatically marks the real-time received ECG signal through the preset interval time, obtains a plurality of marking points, and performs slice processing on the ECG signal through the number of marking points; compared with the sampling mode through the interval time, the acquisition of the signal segment is more flexible and convenient, and the matching between the time sequence information and the signal segment is facilitated by the marking points;

[0053] 3. In order to more accurately determine whether the signal segment has an abnormality, the interference adjustment model identifies a plurality of time sequence information continuous signal segments at a time, judges the deviation wave band by image comparison of the continuous waveform change, calculates the change coefficient of the waveform in the wave band through the construction of the coordinate system, that is, the degree of deviation, and judges whether it belongs to the normal fluctuation range or has an interference through the degree of deviation;

[0054] 4. Through the overall waveform image, the position of the deviation wave band in the overall waveform image and the waveform itself, the interference type matched with the current deviation wave band and the position of the deviation wave band can be screened out from the preset database, and the preset adjustment strategy is correspondingly screened out to adjust the error of the deviation wave band. Error adjustment also includes different adjustment methods, such as baseline calibration, high-frequency interference signal filtering, waveform adjustment, etc., and finally generates a denoised electrocardiogram to the user end. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 is an implementation flowchart of an embodiment of the electrocardiogram anti-interference detection method based on big data analysis of the present application;

[0056] Figure 2 is an implementation flowchart of step S20 in an embodiment of the electrocardiogram anti-interference detection method based on big data analysis of the present application;

[0057] Figure 3is an implementation flowchart of step S32 in an embodiment of an electrocardiogram anti-interference detection method based on big data analysis of the present application;

[0058] Figure 4 is an implementation flowchart of step S50 in an embodiment of an electrocardiogram anti-interference detection method based on big data analysis of the present application;

[0059] Figure 5 is a principle block diagram of a computer device of the present application. DETAILED DESCRIPTION

[0060] The following will be described in combination with the accompanying Figures 1-5 The present application will be further described in detail.

[0061] In embodiments, as shown in Figure 1 The present application discloses an electrocardiogram anti-interference detection method based on big data analysis, which specifically comprises the following steps:

[0062] S10: Real-time receiving of ECG signals output by an electrocardiogram monitor and inputting to a pre-trained interference adjustment model;

[0063] In the embodiment, the ECG signal reflects the contraction and diastolic process of the heart by recording the changes of the electrical activity of the heart, and the sequence of the ECG signal waveform is P wave, PR interval, QRS complex, ST segment, T wave and QT interval; the interference adjustment model is used to identify the interference waveform in the ECG signal and adjust and filter the interference waveform in the ECG signal, which extracts features and establishes a time sequence model through CNN and LSTM, uses dilated causal convolution to process long time sequence dependence, combines course learning, multi-task learning and learning forward training strategy, and constrains the waveform difference between the interference signal and the normal signal through mean square error.

[0064] Specifically, the ECG signal is acquired in real time by the electrocardiogram monitor and inputted to the interference adjustment model in real time.

[0065] S20: The interference adjustment model extracts the ECG signal and generates signal segments of different time sequences based on a preset sampling rule;

[0066] In the embodiment, the sampling rule is the logic of ECG signal sampling, including interval sampling point marking and segment sampling, and the sampling sequence is according to the generation sequence of the ECG signal, i.e. time sequence information, and the signal segment refers to a waveform segment sampled according to the time sequence.

[0067] Specifically, the interference adjustment model samples the ECG signal according to the preset sampling logic, and extracts multiple signal segments according to the generation sequence of the ECG signal.

[0068] S30: The interference adjustment model performs feature recognition on the plurality of signal segments, judges whether there is an interfered signal segment, and marks it as an abnormal segment;

[0069] In this embodiment, the feature recognition is image recognition of the ECG signal waveform features and waveform trends in the signal segment, and the abnormal segment refers to a signal segment with interference signals, including power frequency interference, baseline drift, electromyographic interference, and partial high-frequency noise interference and other types of waveform features.

[0070] Specifically, the interference adjustment model performs waveform feature recognition on the ECG signals of the plurality of signal segments, judges the signal segment with interference signals, and determines the signal segment as an abnormal segment by assigning a preset identifier to the signal segment.

[0071] S40: When there is an abnormal segment, the abnormal segment and other signal segments associated with the time sequence of the abnormal segment are obtained to form a feature set;

[0072] In this embodiment, the time sequence correlation refers to other signal segments extracted within a preset time before and after the abnormal segment, for example, 3 seconds before and 3 seconds after; the feature set includes the obtained signal segments and is arranged according to the sampling time sequence information.

[0073] Specifically, when there is an abnormal segment, in order to clearly analyze the abnormal reasons of the abnormal segment, the adjacent other signal segments associated with the time sequence of the abnormal segment are collected and packaged together to form a feature set.

[0074] S50: Identify and judge the interference type of the feature set, and select the corresponding adjustment strategy based on the identified interference type, perform error adjustment on the signal segments that need to be adjusted in the feature set based on the adjustment strategy, and generate an electrocardiogram and send it to the user end.

[0075] In this embodiment, the interference type is artificially set, and the interference type does not include normal ECG signal types and pathological ECG signal types, and there is one or more than one interference type in the same feature set; the adjustment strategy is a strategy for screening, calibrating, and adjusting the interfered ECG signal waveform.

[0076] Specifically, after identifying and judging the interference type existing in the current feature set, the corresponding adjustment strategy is retrieved based on the interference type, which is used to screen, calibrate, and adjust the interference waveform of the ECG signal in the signal segment, so as to process the interference signal in the abnormal segment, obtain an electrocardiogram that can correctly reflect the ECG signal of the patient, and send it to the user end. The user end is a PC end for the staff to use.

[0077] In an embodiment, with reference to Figure 2 , step S20 includes the following steps:

[0078] S21: The interference adjustment model labels the received ECG signal based on a preset interval duration, to obtain a plurality of label points;

[0079] S22: The ECG signal between adjacent n label points is sampled and extracted to generate a signal segment, n being a positive integer;

[0080] S23: Time information of the n label points in the signal segment is obtained as timing information associated with the signal segment.

[0081] In this embodiment, based on the arrangement rule of different wave bands in the ECG signal, the signal segment is distinguished by setting labels at intervals, the interval duration is a human-set time, which facilitates the interference adjustment model to identify the interference signal in the signal segment, and the matching of the timing information of each label point facilitates the identification and collection of the feature set.

[0082] In an embodiment, step S30 includes the following steps:

[0083] S31: The interference adjustment model performs feature identification on a preset number of signal segments at a time according to the generation order of the signal segments;

[0084] S32: Based on image feature comparison, deviation wave bands existing in a plurality of signal segments are identified and labeled, and a waveform variation coefficient of the deviation wave bands is calculated to determine whether an abnormal segment exists in the plurality of signal segments.

[0085] In this embodiment, the interference adjustment model performs image feature identification on a preset number of signal segments at a time, and the signal segments identified at a time continue in the generation order, i.e., a plurality of adjacent signal segments based on the timing information. The deviation wave bands existing in the signal segments are identified and labeled through image comparison, wherein the judgment of the deviation wave bands is not only based on the feature comparison of the waveforms in the images, but also needs to judge the specific wave band of the signal segment where the deviation wave band exists, because the deviation wave band may exist in the signal segment, which belongs to normal ECG signal or pathological ECG signal waveform.

[0086] After further identification and determination of the deviation wave band, the waveform variation coefficient of the deviation wave band is calculated, i.e., a parameter representing the degree of waveform deviation is calculated, including slope deviation, radian deviation, and abnormal occurrence compared with normal ECG signal and pathological ECG signal; and finally it is determined whether an abnormal segment exists in the plurality of signal deviations;

[0087] If there is no abnormal segment, a preset number of signal segments are obtained again according to the timing information for identification.

[0088] In an embodiment, with reference to Figure 3 , step S32 includes the following steps:

[0089] S321: Obtain an overall waveform image formed by the plurality of signal segments;

[0090] S322: Compare the overall waveform image with a pre-stored reference image in terms of image features through a preset plane coordinate system, and mark a deviation waveband existing with the reference image;

[0091] S323: Obtain coordinate points of the deviation waveband, and calculate a waveform variation coefficient of the deviation waveband;

[0092] S324: Identify a position of the deviation waveband in the overall waveform image, so as to determine a waveband type and a signal segment to which the deviation waveband belongs;

[0093] S325: Match a preset coefficient threshold value based on the waveband type, and compare the coefficient threshold value with the currently calculated waveform variation coefficient, when the waveform variation coefficient is greater than or equal to the coefficient threshold value, it is determined that the signal segment to which the current deviation waveband belongs is an abnormal segment, when the waveform variation coefficient is less than the coefficient threshold value, it is determined that the current deviation waveband is a normal fluctuation range.

[0094] In the embodiment, the position of the waveform can be obtained by combining the plurality of signal segments into the overall waveform image again, and the waveform is further compared with the reference image, so as to obtain a waveband name of the signal segment, thereby accurately determining whether the waveform is a deviation waveform. The reference image includes a waveform image of a normal ECG signal and a waveform of a pathological ECG image.

[0095] The coordinate of the waveform end point and the inflection point of the deviation waveband can be obtained through the plane coordinate system, so that the waveform variation coefficient is calculated, and the waveform type and the signal segment in which the deviation waveband is located are further identified and determined. The waveband type represents the name of the waveband.

[0096] Specifically, the plurality of time sequence related signal segments of the obtained signal are combined into an overall waveform image, the deviation waveband different from the reference image is identified through the overall waveform image, the coefficient threshold value is compared with the waveform variation coefficient based on the deviation variation coefficient of the deviation waveband, the signal segment and the waveband type, so as to determine the deviation degree of the current deviation waveband, thereby finally determining whether there is an abnormal segment that needs to be adjusted.

[0097] In an embodiment, referring to Figure 4 , step S50 includes the following steps:

[0098] S51: Match an interference type of the current feature set based on the overall waveform image and the deviation waveband in the feature set, and the number of the interference types includes one or more than one;

[0099] S52: filtering and / or calibrating the abnormal segment based on the corresponding adjustment strategy selected according to the interference type;

[0100] S53: combining the signal segments after error adjustment according to the time sequence information to generate an electrocardiogram and sending the electrocardiogram to a user terminal.

[0101] In the embodiment, the determination of the interference type needs to combine the overall waveform image and the position relationship of the deviation band in the image; further, the overall waveform image after combination is restored to a plurality of signal segments, and the filtering and / or calibration of the deviation band in the abnormal segment by adjustment can reduce the data amount during adjustment strategy processing.

[0102] In an embodiment, step S51 includes the following steps:

[0103] S511: identifying the distribution position of the deviation band in the overall waveform image based on the context association rule; identifying the band type and the waveform change coefficient of the deviation band one by one;

[0104] S512: matching the interference type of the current deviation band one by one from the preset interference type database as the interference type of the current feature set.

[0105] In the embodiment, the context association relationship is used to determine the probability of the occurrence of the deviation band in the current band type and the general size of the change coefficient of the occurrence, so as to determine whether the deviation band is interference, and the interference type of the deviation band is identified one by one. In other embodiments, a plurality of deviation bands can be identified at the same time through specific GPU distribution, so as to improve the efficiency of abnormal segment identification.

[0106] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the application.

[0107] In an embodiment, an electrocardiogram anti-interference detection system based on big data analysis is provided, which corresponds to the electrocardiogram anti-interference detection method based on big data analysis in the above embodiment. The electrocardiogram anti-interference detection system based on big data analysis includes:

[0108] A signal receiving module is configured to receive an ECG signal output by an electrocardiogram monitor in real time and input the ECG signal into a pre-trained interference adjustment model;

[0109] A sampling module is configured to extract the ECG signal and generate signal segments at different time sequences based on a preset sampling rule.

[0110] The identification module is used by the interference adjustment model to identify features of multiple signal segments, determine whether there are any interfered signal segments, and mark them as abnormal segments;

[0111] The association module is used to acquire the abnormal segment and other signal segments that are time-series associated with the abnormal segment when an abnormal segment exists, and form a feature set;

[0112] The adjustment module is used to identify and determine the interference type of the feature set, and select the corresponding adjustment strategy based on the identified interference type. Based on the adjustment strategy, the signal segments in the feature set that need to be adjusted are adjusted for error, and an electrocardiogram is generated and sent to the user terminal.

[0113] Optionally, the sampling module includes:

[0114] The marking submodule is used by the interference adjustment model to mark the received ECG signal based on a preset interval, resulting in multiple marking points;

[0115] The sampling submodule is used to sample and extract the ECG signal between n adjacent marker points to generate a signal segment, where n is a positive integer;

[0116] The timing association submodule is used to obtain the timing information of n marker points in a signal segment, which is used as the timing information associated with the signal segment.

[0117] Optionally, the recognition module includes:

[0118] The feature recognition submodule is used to identify the features of a preset number of signal segments in a single operation based on the generation order of the signal segments in the interference adjustment model.

[0119] The deviation calculation submodule is used to identify and mark the deviation bands in several signal segments based on image feature comparison, and calculate the waveform change coefficient of the deviation bands to determine whether there are abnormal segments in the current several signal segments.

[0120] Optional, the deviation calculation submodule includes:

[0121] The waveform image acquisition unit is used to acquire the overall waveform image formed by several signal segments;

[0122] The deviation band marking unit is used to compare the overall waveform image with the pre-stored reference image using a preset planar coordinate system, and mark the deviation bands that exist with the reference image.

[0123] The coefficient calculation unit is used to obtain the coordinate points of the deviation band and calculate the waveform change coefficient of the band where the deviation occurs.

[0124] The identification unit is configured to identify the position of the deviating wave band in the overall waveform image, so as to determine the wave band type and the signal segment to which the deviating wave band belongs.

[0125] The judgment unit is configured to match a preset coefficient threshold based on the wave band type, and compare the waveform change coefficient currently calculated with the coefficient threshold, when the waveform change coefficient is greater than or equal to the coefficient threshold, it is determined that the signal segment to which the current deviating wave band belongs is an abnormal segment, and when the waveform change coefficient is less than the coefficient threshold, it is determined that the current deviating wave band is within a normal fluctuation range.

[0126] Optionally, the adjustment module comprises:

[0127] The interference judgment sub-module is configured to match the interference type of the current feature set based on the overall waveform image and the deviating wave band in the feature set, and the number of interference types includes one or more types.

[0128] The adjustment sub-module is configured to filter and / or calibrate the interference signal based on the interference type to adjust the error of the deviating wave band in the abnormal segment.

[0129] The generation sub-module is configured to combine the signal segment after error adjustment according to the time sequence information to generate an electrocardiogram and send it to the user end.

[0130] Optionally, the interference judgment sub-module comprises:

[0131] The judgment unit is configured to identify the distribution position of the deviating wave band in the overall waveform image based on the context association rule, and identify the wave band type and the waveform change coefficient of the deviating wave band one by one.

[0132] The matching unit is configured to match the interference type of the current deviating wave band one by one from the preset interference type database, as the interference type of the current feature set.

[0133] The specific limitation of the electrocardiogram anti-interference detection system based on big data analysis can be referred to the limitation of the electrocardiogram anti-interference detection method based on big data analysis in the above, which will not be repeated here. Each module in the electrocardiogram anti-interference detection system based on big data analysis can be realized by software, hardware and their combination. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so that the processor can call and execute the operation corresponding to each module.

[0134] In one embodiment, a computer device is provided, which can be a server, and its internal structure diagram can be as shown in Figure 5As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the computer device is used to communicate with the external terminal through the network connection. The computer program is executed by the processor to implement an electrocardiogram anti-interference detection method based on big data analysis.

[0135] In one embodiment, a computer device is provided, including a memory, a processor and a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement an electrocardiogram anti-interference detection method based on big data analysis.

[0136] In one embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement an electrocardiogram anti-interference detection method based on big data analysis.

[0137] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Among them, any reference to memory, storage, database or other medium used in each embodiment provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM) and the like.

[0138] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

[0139] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A big data analysis-based electrocardiogram anti-interference detection method, characterized in that: real-time receiving an ECG signal output by an electrocardiogram monitor and inputting the ECG signal into a pre-trained interference adjustment model; the interference adjustment model extracts the ECG signal and generates signal segments of different time sequences based on a preset sampling rule; the interference adjustment model performs feature recognition on a plurality of signal segments, judges whether there is a signal segment disturbed by interference, and marks the signal segment as an abnormal segment; when there is an abnormal segment, obtaining the abnormal segment and other signal segments associated with the abnormal segment in time sequence, forming a feature set; identifying and judging the interference type of the feature set, and screening out a corresponding adjustment strategy based on the identified interference type, performing error adjustment on the signal segments that need to be adjusted in the feature set based on the adjustment strategy, and generating an electrocardiogram and sending the electrocardiogram to a user terminal; the step of the interference adjustment model performing feature recognition on a plurality of signal segments, judging whether there is a signal segment disturbed by interference, and marking the signal segment as an abnormal segment, comprises: the interference adjustment model performs feature recognition on a preset number of signal segments at a time according to the generation order of the signal segments; based on image feature comparison, identifying and marking the deviation wave band existing in the plurality of signal segments, calculating the waveform variation coefficient of the deviation wave band, and judging whether there is an abnormal segment in the plurality of signal segments; in the step of based on image feature comparison, identifying and marking the deviation wave band existing in the plurality of signal segments, calculating the waveform variation coefficient of the deviation wave band, and judging whether there is an abnormal segment in the plurality of signal segments, comprising: obtaining an overall waveform image formed by the plurality of signal segments; comparing the overall waveform image with a pre-stored reference image in image features through a preset plane coordinate system, and marking the deviation wave band existing in the reference image; obtaining the coordinate points of the deviation wave band, and calculating the waveform variation coefficient of the deviation wave band; identifying the position of the deviation wave band in the overall waveform image to determine the wave band type and the signal segment to which the deviation wave band belongs; matching the wave band type with a preset coefficient threshold value, and comparing the calculated waveform variation coefficient with the coefficient threshold value, when the waveform variation coefficient is greater than or equal to the coefficient threshold value, determining that the signal segment to which the current deviation wave band belongs is an abnormal segment, and when the waveform variation coefficient is less than the coefficient threshold value, determining that the current deviation wave band is within a normal fluctuation range; the feature set includes the signal segments other than the abnormal segment identified by the interference adjustment model at a time; by combining the plurality of signal segments into an overall waveform image again, the position of the waveform can be obtained, and the waveform is further compared with the reference image to obtain the wave band name of the signal segment, so as to accurately judge whether the waveform is a deviation waveform; the reference image includes the waveform image of a normal ECG signal and the waveform of a pathological ECG image; through the plane coordinate system, the coordinates of the waveform endpoints and inflection points of the deviation wave band can be obtained, so as to calculate the waveform variation coefficient and further identify and determine the waveform type and the signal segment to which the deviation wave band belongs; the wave band type represents the name of the wave band. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ 2. The electrocardiogram anti-interference detection method based on big data analysis according to claim 1, characterized in that, The interference adjustment model extracts the ECG signal and generates signal segments of different time sequences based on a preset sampling rule, and the steps include: The interference adjustment model labels the received ECG signal based on a preset interval length to obtain a plurality of marking points; The ECG signal between adjacent n marking points is sampled and extracted to generate a signal segment, and n is a positive integer; Obtain the time information of the n marking points in the signal segment as the time sequence information associated with the signal segment.

3. The electrocardiogram anti-interference detection method based on big data analysis according to claim 1, characterized in that, The steps of identifying and judging the interference type of the feature set, and selecting the corresponding adjustment strategy based on the identified interference type, adjusting the error of the signal segment that needs to be adjusted in the feature set based on the adjustment strategy, and generating an electrocardiogram and sending it to the user end include: Based on the overall waveform image and the deviation wave band in the feature set, the interference type of the current feature set is matched, and the number of interference types includes one or more; Based on the interference type, the adjustment strategy is selected to adjust the error of the deviation wave band in the abnormal segment, and the error adjustment includes filtering and / or calibration processing of the interference signal; The signal segment after error adjustment is combined according to the time sequence information to generate an electrocardiogram and send it to the user end.

4. The electrocardiogram anti-interference detection method based on big data analysis according to claim 3, characterized in that, The steps of matching the interference type of the current feature set based on the overall waveform image and the deviation wave band in the feature set include: Based on the context association rule, identify the distribution position of the deviation wave band in the overall waveform image; identify the wave band type and waveform change coefficient of the deviation wave band one by one; Match the interference type of the current deviation wave band one by one from the preset interference type database as the interference type of the current feature set.

5. An electrocardiogram anti-interference detection system based on big data analysis, for implementing the electrocardiogram anti-interference detection method based on big data analysis as claimed in any one of claims 1 to 4, characterized in that: a signal receiving module for receiving the ECG signal output by the electrocardiogram monitor in real time and inputting it into the interference adjustment model that has been trained in advance; a sampling module for the interference adjustment model to extract the ECG signal and generate signal segments of different time sequences based on a preset sampling rule; an identification module for the interference adjustment model to perform feature identification on a plurality of signal segments, judge whether there is a signal segment disturbed by interference, and mark it as an abnormal segment; an association module for obtaining the abnormal segment and other signal segments associated with the time sequence of the abnormal segment when the abnormal segment exists, and forming a feature set; an adjustment module for identifying and judging the interference type of the feature set, and selecting the corresponding adjustment strategy based on the identified interference type, adjusting the error of the signal segment that needs to be adjusted in the feature set based on the adjustment strategy, and generating an electrocardiogram and sending it to the user end.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the electrocardiogram anti-interference detection method based on big data analysis as claimed in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, wherein the computer program comprises the following steps of: receiving a request for a resource from a client; determining whether the client is authorized to access the resource; and if the client is authorized to access the resource, providing the resource to the client. The computer program is executed by the processor to implement the steps of the electrocardiogram anti-interference detection method based on big data analysis as claimed in any one of claims 1 to 4.

Images