A magnetocardiogram analysis method and device, electronic equipment, and storage medium

Abstract

The present application provides a magnetocardiogram analysis method and device, electronic device, and storage medium. The method includes: obtaining a magnetocardiogram data set to be analyzed, and segmenting at least one band of magnetocardiogram data sets from the magnetocardiogram data set to be analyzed; determining a first set of angles of a target vector corresponding to the band based on the magnetocardiogram data set of each band, and determining the jump wavelet range of the band based on a preset angle fluctuation amplitude threshold; for each band, determining the angles that meet the jump wavelet range of the band as continuous angles, and other angles as jump angles, and determining multiple jump features based on the characteristics of the continuous angles and the characteristics of the jump angles; and calculating the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram based on the multiple jump features to obtain the analysis result of the magnetocardiogram to be analyzed. The present application obtains a more accurate analysis result by extracting the jump features of the target vector in the magnetocardiogram, and performing quantitative analysis and calculation based on the jump features.

Description

Technical Field

[0001] The present application relates to the technical field of biomedical signal analysis, and in particular to a magnetocardiogram analysis method and device, electronic equipment, and storage medium. Background Art

[0002] Magnetocardiography is a cardiac function test that detects the weak magnetic field generated by the heart's electrical activity. When the heart is abnormal, the magnetic field changes, and these changes can be accurately detected by the magnetocardiograph.

[0003] The biggest feature of the magnetocardiogram is the dipole, including positive and negative poles, whose relative positions change continuously throughout the entire cycle of cardiac electrical activity and show different changing patterns in different bands. The existing method mainly extracts the dipole distance and dipole pointing angle, and then qualitatively analyzes the changing patterns of different bands of the magnetocardiogram by manually viewing the magnetocardiogram and the changes in these two features. However, such a qualitative analysis method lacks standardized analysis rules and cannot accurately describe the overall changing patterns of the bands, so it is not accurate enough. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a magnetocardiogram analysis method and device, electronic device, and storage medium, by extracting the jump features of the target vector in the magnetocardiogram, and performing quantitative analysis and calculation based on the jump features to obtain more accurate analysis results.

[0005] This application mainly includes the following aspects:

[0006] In a first aspect, an embodiment of the present application provides a magnetocardiogram analysis method, the method comprising:

[0007] Acquire a magnetocardiogram data set to be analyzed, and segment the magnetocardiogram data set into at least one band from the magnetocardiogram data set to be analyzed;

[0008] Based on the magnetocardiogram data set of each band, a first rotation angle set of the rotation angle of the target vector corresponding to the band is determined; wherein the target vector is a vector formed by the magnetocardiogram dipoles in the isomagnetic map corresponding to the magnetocardiogram to be analyzed or a vector corresponding to the main current in the current density map corresponding to the magnetocardiogram to be analyzed; the rotation angle of the target vector is the rotation angle of the target vector between two adjacent frames of isomagnetic maps or current density maps;

[0009] For each band of the magnetocardiogram data set, based on a first rotation angle set of a target vector corresponding to the band and a preset rotation angle fluctuation amplitude threshold, a jump wavelet range of the band is determined; the jump wavelet range of the band is an angle range of the target vector corresponding to the band in a continuous rotation process;

[0010] For each band of the magnetocardiogram data set, determining the corners that meet the jump wavelet range of the band as continuous corners, and other corners as jump corners, and determining a plurality of jump features according to the features of the continuous corners and the features of the jump corners;

[0011] The similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram is calculated according to the multiple jump features, and the analysis result of the magnetocardiogram to be analyzed is obtained according to the similarity.

[0012] In a second aspect, an embodiment of the present application further provides a magnetocardiogram analysis device, the magnetocardiogram analysis device comprising:

[0013] A data acquisition module, used for acquiring a magnetocardiogram data set to be analyzed, and segmenting the magnetocardiogram data set into at least one band from the magnetocardiogram data set to be analyzed;

[0014] The turning angle set determination module is used to determine the first turning angle set of the turning angle of the target vector corresponding to the band based on the magnetocardiogram data set of each band respectively; wherein the target vector is the vector formed by the magnetocardiogram dipoles in the isomagnetic map corresponding to the magnetocardiogram to be analyzed or the vector corresponding to the main current in the current density map corresponding to the magnetocardiogram to be analyzed; the turning angle of the target vector is the turning angle of the target vector between two adjacent frames of isomagnetic maps or current density maps;

[0015] A jump wavelet range determination module is used to determine the jump wavelet range of each band of the magnetocardiogram data set based on a first rotation angle set of the rotation angle of the target vector corresponding to the band and a preset rotation angle fluctuation amplitude threshold; the jump wavelet range of the band is the rotation angle range of the target vector corresponding to the band in the continuous rotation process;

[0016] A jump feature extraction module is used to determine, for each band of the magnetocardiogram data set, that the corners that meet the jump wavelet range of the band are continuous corners, and other corners are jump corners, and determine multiple jump features according to the characteristics of the continuous corners and the characteristics of the jump corners;

[0017] The image analysis module is used to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the multiple jump features, and obtain the analysis result of the magnetocardiogram to be analyzed according to the similarity.

[0018] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the magnetocardiogram analysis method as described above are performed.

[0019] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the magnetocardiogram analysis method described above are executed.

[0020] The magnetocardiogram analysis method and device, electronic device, and storage medium provided in the embodiments of the present application first use a magnetocardiogram instrument to collect a magnetocardiogram data set to be analyzed, and divide the magnetocardiogram data set into at least one band from the data set; then perform feature extraction on the magnetocardiogram data set of the band to obtain a target vector jump feature of the band; finally, use the target vector jump feature to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram, and analyze the magnetocardiogram to be analyzed based on the calculation result.

[0021] Compared with the qualitative analysis method of magnetocardiogram in the prior art, the present application extracts the jump characteristic parameters within the divided band based on the jump characteristics of the dipole or main current, which can quantitatively and accurately display the electrophysiological changes of the heart, thereby improving the analysis capability of magnetocardiogram and the accuracy of the analysis results.

[0022] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 A flowchart of a magnetocardiogram analysis method provided in an embodiment of the present application is shown;

[0025] Figure 2 A flowchart of another magnetocardiogram analysis method provided in an embodiment of the present application is shown;

[0026] Figure 3 A schematic diagram of a segmented vector area based on pole distance provided in an embodiment of the present application is shown;

[0027] Figure 4 A schematic diagram of segmented vector area based on amplitude provided in an embodiment of the present application is shown;

[0028] Figure 5 A schematic diagram of a segmented average vector area based on pole distance provided in an embodiment of the present application is shown;

[0029] Figure 6 A schematic diagram of a segmented average vector area based on amplitude provided in an embodiment of the present application is shown;

[0030] Figure 7 A flowchart of another magnetocardiogram analysis method provided in an embodiment of the present application is shown;

[0031] Figure 8 A flowchart of another magnetocardiogram analysis method provided in an embodiment of the present application is shown;

[0032] Fig. 9 A schematic diagram of the structure of a magnetocardiogram analysis device provided in an embodiment of the present application is shown;

[0033] Fig.10 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown.

[0034] In the figure:

[0035] 910 -data acquisition module, 920 -angle set determination module, 930 -jump wavelet range determination module, 940 -jump feature extraction module, 950 -image analysis module, 1000 -electronic device, 1010 -processor, 1020 -memory, 1030 -bus. DETAILED DESCRIPTION

[0036] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0037] First, the application scenarios to which this application is applicable are introduced. This application uses non-invasive magnetocardiography technology to measure the weak magnetic field generated by the heart to evaluate the electrical activity of the heart, and can be applied in the fields of medicine and bioengineering.

[0038] Research has found that the existing magnetocardiogram analysis method usually involves manually viewing the magnetocardiogram, analyzing the changes in dipole distance and dipole vector angle, and using qualitative analysis methods to obtain the magnetocardiogram change law. However, the analysis results obtained in this way are often subjective and cannot accurately describe the overall change law of the band, and have low accuracy.

[0039] Based on this, an embodiment of the present application provides a magnetocardiogram analysis method, which obtains a more accurate magnetocardiogram analysis result by quantitatively analyzing and calculating the jump characteristics of the magnetocardiogram band.

[0040] See also Figure 1 , Figure 1 This is a flow chart of a magnetocardiogram analysis method provided in an embodiment of the present application. Figure 1 As shown in , the magnetocardiogram analysis method provided in the embodiment of the present application includes:

[0041] S101, obtaining a magnetocardiogram dataset to be analyzed, and segmenting the magnetocardiogram dataset to be analyzed into at least one band of magnetocardiogram dataset.

[0042] In this step, firstly, a magnetocardiogram data set to be analyzed is obtained, which is a data signal obtained by a multi-channel magnetocardiogram acquisition instrument within a period of time, and generally may include magnetic field data, timestamp data, etc. It is understandable that, since the magnetocardiogram data set to be analyzed is data within a period of time, the magnetocardiogram data set to be analyzed may be divided into magnetocardiogram data sets of multiple bands based on the cardiac cycle, such as P band, QRS band, ST band, T band, etc.

[0043] S102. Based on the magnetocardiogram data set of each band, determine a first set of rotation angles of a target vector corresponding to the band; wherein the target vector is a vector formed by magnetocardiogram dipoles in an isomagnetic map corresponding to the magnetocardiogram to be analyzed or a vector corresponding to a main current in a current density map corresponding to the magnetocardiogram to be analyzed; and the rotation angle of the target vector is an angle of the target vector between two adjacent frames of isomagnetic maps or current density maps.

[0044] In this step, the set of rotation angles of the target vector is determined for each band of the magnetocardiogram dataset, and then the band is analyzed based on the set of rotation angles in the subsequent steps. Specifically, features are extracted from the magnetocardiogram dataset of the band to determine the change in the angle of the target vector, that is, the rotation angle of the target vector. Among them, the target vector is a vector with a jump rotation phenomenon, such as the vector formed by the central magnetic dipole of the isomagnetic map, the vector corresponding to the main current in the current density map, etc. The isomagnetic map or current density map can be drawn based on the magnetocardiogram dataset to be analyzed, and a frame of isomagnetic map or current density map is obtained at each moment.

[0045] S103. For each band of the magnetocardiogram dataset, based on a first angle set of the angle of the target vector corresponding to the band and a preset angle fluctuation amplitude threshold, determine the jump wavelet range of the band; the jump wavelet range of the band is the angle range of the target vector corresponding to the band in the continuous rotation process.

[0046] In this step, after obtaining the set of angles of the target vector, it is considered that the change of the angle of the target vector has a certain regularity, for example, jumps often occur only a small number of times, and the angle during jumps is generally large, while the angle during continuous rotation is generally small. Therefore, based on this feature, the maximum continuous rotation angle except for jumps can be used to limit the jump wavelet range wg. The angle within this range is the angle during continuous rotation, and the angle outside this range is the angle during jumps. For the convenience of calculation, in the specific range determination process, the volatility of the angle can be used as the basis for range determination. If the volatility of the angle is large, it is considered that a jump has occurred, otherwise it is considered that continuous rotation has occurred.

[0047] Specifically, according to the volatility of the corner, it is determined that the corner with a small fluctuation amplitude belongs to the jump wavelet range, and based on this, the boundary of the jump wavelet range is limited, that is, the preset corner fluctuation amplitude threshold. Specifically, the variance, standard deviation, etc. can be used as the measurement standard of the corner fluctuation amplitude. Finally, the corner fluctuation amplitude is small, that is, the amplitude less than the preset corner fluctuation amplitude threshold is taken as the small fluctuation amplitude.

[0048] S104. For each band of the magnetocardiogram data set, determine the corners that meet the jump wavelet range of the band as continuous corners, and other corners as jump corners, and determine multiple jump features according to the features of the continuous corners and the features of the jump corners.

[0049] In this step, for each band of the magnetocardiogram data set, each specific jump is compared with the jump wavelet range to obtain the corresponding jump feature. Specifically, in each band of the magnetocardiogram data set, according to the jump wavelet range, the turning angle is divided into continuous turning angle and jump turning angle, and the characteristics of the continuous turning angle and the jump turning angle, such as the angle and angular velocity of the turning angle, are determined respectively, and then the characteristics of the continuous turning angle and the jump turning angle are comprehensively analyzed to obtain the jump feature for the band.

[0050] S105 , calculating the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the multiple jump features, and obtaining an analysis result of the magnetocardiogram to be analyzed according to the similarity.

[0051] In this step, the similarity between the jump characteristics of the magnetocardiogram to be analyzed and the jump characteristics of the preset target magnetocardiogram is calculated. The higher the similarity, the more similar the cardiac electrical activity reflected by the magnetocardiogram to be analyzed is to the reference. Therefore, the magnetocardiogram analysis can be performed with the help of the target magnetocardiogram to obtain the corresponding results. For example, a magnetocardiogram in a healthy state can be used as a target magnetocardiogram. If the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram is high, the cardiac electrical activity reflected by the magnetocardiogram is considered normal. Otherwise, it is considered that there may be an abnormality, and further analysis can be performed to determine the type and location of the abnormality.

[0052] Among them, if in the aforementioned steps, only one band is divided and only the jump features of the band are extracted, then in step S105, the similarity needs to be calculated for the jump features of the band; if in the aforementioned steps, multiple bands are divided and the jump features of each band are extracted respectively, then in step S105, the similarity needs to be calculated for the jump features of each band respectively, and then the similarity of each band is summarized to obtain the similarity between the images.

[0053] The magnetocardiogram analysis method provided in the embodiment of the present application first uses a magnetocardiogram instrument to collect a magnetocardiogram data set to be analyzed, and divides the magnetocardiogram data set into at least one band from the data set; then, feature extraction is performed on the magnetocardiogram data set of the band to obtain a target vector jump feature of the band; finally, the target vector jump feature is used to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram, and the magnetocardiogram to be analyzed is analyzed according to the calculation result.

[0054] Compared with the qualitative analysis method of magnetocardiogram in the prior art, the present application extracts the jump characteristic parameters within the divided band based on the jump characteristics of the dipole or main current, which can quantitatively and accurately display the electrophysiological changes of the heart. Therefore, the analysis capability of magnetocardiogram is improved and the accuracy of the analysis results is improved.

[0055] Optionally, in step S102, based on the magnetocardiogram data set of each band, a first set of rotation angles of the target vector corresponding to the band is determined, including:

[0056] S102-a, based on the magnetocardiogram data set of each band, draw the isomagnetic map of the band, and determine the vector angle of the magnetocardiogram in each frame of the isomagnetic map according to the magnetic field strength of each frame of the isomagnetic map, obtain the first set of rotation angles of the magnetocardiogram between every two frames of the isomagnetic map, and determine the first set as the first rotation angle set of the target vector corresponding to the band; or,

[0057] S102-b, based on the magnetocardiogram data set of each band, draw the current density map of the band, and determine the second set of the main current angles between every two frames of the current density map according to the main current direction in each frame of the current density map, and determine the second set as the first angle set of the target vector corresponding to the band.

[0058] In this step, several methods for determining the rotation angle of the target vector are provided. Specifically, for the dipole vector, a rotation angle determination method in step S102-a is provided; for the main current, a rotation angle determination method in step S102-b is provided.

[0059] Specifically, for step S102-a, first draw an isomagnetic map based on the magnetocardiogram data set, wherein each frame of the magnetocardiogram corresponds to one frame of the isomagnetic map. Then, based on the magnetic field strength of each frame of the isomagnetic map, determine the position coordinates of the positive pole at the maximum magnetic field strength and the position coordinates of the negative pole at the minimum magnetic field strength. The direction in which the positive pole points to the negative pole is the dipole vector direction, from which the dipole vector angle can be obtained. Finally, determine the rotation angle of the magnetocardiogram dipole based on the change in the vector angle in each adjacent frame of the isomagnetic map. Among them, if the vector direction of the latter frame rotates counterclockwise compared to the previous frame, the rotation angle is determined to be a negative value; otherwise, the rotation angle is determined to be a positive value, and the final rotation angle value is (-180°, 180°].

[0060] For step S102-b, a current density map is first drawn based on the magnetocardiogram data set, wherein each frame of the magnetocardiogram corresponds to a frame of the current density map. In the current density map, the current direction is the direction of the current density, so the change of the main current angle, that is, the turning angle of the main current between two frames of the current density map, can be determined based on the main current direction of each frame of the current density map.

[0061] See also Figure 2 , Figure 2 This is a flow chart of a magnetocardiogram analysis method provided by another embodiment of the present application. Figure 2 As shown in , in step 102, the preset turning angle fluctuation amplitude threshold includes a mean square error threshold. For each band of the magnetocardiogram data set, the set of turning angles of the target vector corresponding to the band and the preset turning angle fluctuation amplitude threshold are used to determine the jump wavelet range of the band, including:

[0062] S201. Arrange the corners in the first corner set from large to small according to their absolute values ​​to obtain a second corner set.

[0063] S202, respectively calculating mean square errors of a preset number of adjacent corners in the second corner set to obtain a mean square error set corresponding to the second corner set.

[0064] S203: Determine the mean square error in the mean square error set that is smaller than a preset turning angle fluctuation amplitude threshold as a small fluctuation mean square error.

[0065] S204, determining the corner in the second corner set corresponding to the small fluctuation mean square error as the small fluctuation corner in the second corner set, and determining the jump wavelet range according to the maximum value of the small fluctuation angle.

[0066] In steps S202-S204, according to the volatility of the corner, it is determined that the corner with a small fluctuation amplitude belongs to the jump wavelet range, and the boundary of the jump wavelet range is limited based on this. For example, first, for the dipole vector angles of all frames of the QR segment, the corners between each two adjacent frames are taken to form an angle set R, and then the elements in the set R are sorted from large to small according to the absolute value to obtain a sequence R1. At this time, for each element in the sequence, the element adjacent to it is the element with the smallest gap with this element.

[0067] Then, for the above sequence R1, take the three adjacent elements to calculate the mean square error, and use the calculated mean square error as the fluctuation range of these three elements. The smaller the mean square error, the smaller the fluctuation range, so the mean square error records whether the angle belongs to the small fluctuation range, that is, whether the angle exists in the continuous rotation process. Of course, the mean square error can also be calculated by taking four adjacent elements, which is not limited here.

[0068] Finally, for the above mean square error, a mean square error set M is obtained, and each mean square error in the set M is judged in turn whether it is less than the corner fluctuation amplitude threshold m0. If it is less than, the fluctuation amplitude corresponding to the mean square error is considered to be a small fluctuation amplitude. Find the mean square error that is less than the corner fluctuation amplitude threshold m0, and then infer the value of the small fluctuation corner corresponding to the mean square error that is less than the corner fluctuation amplitude threshold m0 in the sequence R1, and record the maximum value of the corner value in each small fluctuation corner as the jump wavelet range wg. Among them, the corner fluctuation amplitude threshold is pre-set and can be determined based on information statistics of a large sample data set and clinical experience.

[0069] For example, for the target vector angles in the following 9 adjacent frames of images: [1, 3, 5, 8, 35, 38, 39, 80, 82], we can get the angle set R: [2, 2, 3, 27, 3, 1, 41, 2]. Sort the elements in the set R from large to small according to the absolute value, and get the angle sequence R1: [41, 27, 3, 3, 2, 2, 2, 1]. Calculate the mean square error of the three adjacent elements in R1 in turn, and get the mean square error set M: [246, 128, 0.22, 0, 0.22]. If the preset angle fluctuation amplitude threshold is 1, the angle fluctuation amplitude corresponding to [0.22, 0, 0.22] in the mean square error set is considered to be a small fluctuation amplitude. It is reversed to the angle value in the sequence R1, and the value of the small fluctuation angle that meets the jump wavelet range wg is obtained: [3, 3, 2, 2, 2, 1], where wg is 3. Therefore, in the aforementioned corner set R: [2, 2, 3, 27, 3, 1, 41, 2], [27, 41] is a jump corner, and [2, 2, 3, 3, 1, 2] is a continuous corner.

[0070] Optionally, the multiple jump features include at least one or more of fluctuation category features, angle category features, area category features and magnetic pole vector category features.

[0071] The fluctuation category feature includes at least one or more of the number of jumps, the jump wavelet range, the maximum clockwise jump angle, and the maximum counterclockwise jump angle. The fluctuation category feature is determined according to the continuous turning angle and the jump turning angle.

[0072] Specifically, the fluctuation category feature is used to characterize the fluctuation amplitude during the rotation process, and specifically includes one or more of the following four features:

[0073] Number of jumps num: The number of corners in this band whose absolute value is greater than the aforementioned wg, that is, the number of jump corners.

[0074] Jump wavelet range wg: obtained according to the above steps, which is the maximum value of the continuous rotation angles.

[0075] Maximum clockwise jump angle pm: the maximum value of the clockwise jump angle among the jump angles of this band. If there is no clockwise jump, pm=0.

[0076] Maximum counterclockwise jump angle nm: the maximum value of the counterclockwise jump angle among the jump angles of this band. If there is no counterclockwise jump, nm=0.

[0077] For example, as in the aforementioned embodiment, wg=3, the jump angle is [27, 41], which are all clockwise angles, so nmu=2, wg=3, pm=41, nm=0.

[0078] The corner classification feature includes at least one or more of the corner sum, angular velocity, corner mean square error, weighted angular velocity and weighted corner mean square error. The corner classification feature is determined according to the continuous corners of each segment in the band, and the segments are divided according to the jump angles in the band.

[0079] Specifically, the angle category feature is used to characterize the rotation angle and angular velocity during the rotation process, and specifically includes one or more of the following five features:

[0080] Corner and r: The jump in the band divides the entire band into several segments, and the corners in each segment are continuous corners. For example, in the aforementioned corner set R: [2, 2, 3, 27, 3, 1, 41, 2], there are two jumps, and the jump corners are 27 and 41 respectively, dividing the entire band into the following three segments: [2, 2, 3], [3, 1] and [2].

[0081] In each segment, the frame-by-frame rotation angle, that is, the union of the continuous rotation angles, is recorded as the rotation angle sum ri of the segment; the rotation angle sum ri of all segments is added together to obtain the rotation angle sum r of the band.

[0082] Angular velocity V: The average angular velocity of each segment in the band is recorded as vi, and the average of vi of all segments is taken to obtain the angular velocity V of the band. For example, the angular velocities of the above three segments are v1, v2 and v3, then

[0083] Mean square error of rotation angle: Similar to the calculation method of angular velocity V, the mean square error of each segment in the band is recorded as msei. The msei of all segments is averaged to obtain the mean square error of rotation angle Mse of the band.

[0084] Weighted angular velocity V1: Since the number of frames in each segment may be different, when taking the average of vi for all segments, we can first weight the average angular velocity vi of each segment according to the number of frames in each segment, and then take the average of all frames to get V1. The more frames a segment has, the higher the weight corresponding to the segment. For example, the number of frames in the above three segments is 4, 3, and 2 respectively, then

[0085] Weighted mean square error of the angle: Similar to the calculation method of the weighted angular velocity V1, first weight the mean square error msei of each segment according to the number of frames of each segment, and then take the average of all frames to obtain Mse1.

[0086] The area category feature includes segment vector area and / or segment average vector area. The area category feature is determined according to the area swept by the continuous corners of each segment.

[0087] Specifically, the area category feature is used to characterize the area swept during the rotation process, and specifically includes at least one of the following two features:

[0088] Segment vector area: In each segment, the area swept by the target vector angle during the rotation process, specifically including the segment vector area ad based on pole pitch and / or the segment vector area af based on amplitude.

[0089] Among them, the calculation method of the segmented vector area ad based on the pole distance is as follows: in each segment, the positive poles are regulated to the same starting point, the dipole vector is used as the angle, the dipole distance is used as the length, and after normalization according to the isomagnetic map dimension, the vector area of ​​the segment is calculated based on the angle and length.

[0090] The segment vector area ad based on the pole pitch may also be calculated by the following method: taking the main current direction as the angle and the main current intensity as the length, the vector area of ​​the segment is calculated based on the angle and the length.

[0091] Among them, the calculation method of the segmented vector area af based on amplitude is as follows: in each segment, the positive pole is regulated to the same starting point, the dipole vector or the main current direction is used as the angle, the magnetic field amplitude is used as the length, and after normalization according to the maximum magnetic field amplitude in the entire band, the vector area of ​​the segment is calculated based on the angle and length.

[0092] See also Figure 3 , Figure 4 , Figure 3 is a schematic diagram of segmented vector area based on pole distance, Figure 4 is a schematic diagram of segmented vector area based on amplitude. Figure 3 As shown, Figure 3 The numbers 1, 2, 3, and 4 in the figure represent four quadrants, respectively. The five sector-shaped shaded areas in the figure are five segments. The blank parts between the segments are obtained by jumping, and the segments are obtained by continuous rotation. For each segment, the dipole vector is used as the angle and the dipole distance is used as the length to obtain the vector area of ​​the segment. Figure 4 As shown, Figure 4 The five sector-shaped shaded areas in the figure are five segments, Figure 3 The five sectors correspond to each other one by one. For each segment, the dipole vector is used as the angle and the current amplitude is used as the length to obtain the vector area of ​​the segment. Figure 3 and Figure 4 The angles are all about the dipole vector, so Figure 3 and Figure 4 The angles of the corresponding sectors in are the same; since Figure 3 and Figure 4 The lengths used in Figure 3 and Figure 4The lengths of the radii of the corresponding sectors are different, and the shapes of the arcs are also different.

[0093] Segmented average vector area: The average of the target vector angles of all frames of each segment is recorded as the average vector of the segment, and the area swept by the average vector of all segments of the entire band during the rotation process is recorded as the segmented average vector area. The segmented average vector area specifically includes the segmented average vector area based on the pole distance and / or the segmented average vector area based on the amplitude, and its calculation method is similar to that of the segmented vector area.

[0094] Among them, the segmented average vector area based on pole pitch is calculated as follows: the average of the dipole vector angles of all frames of each segment is recorded as the average vector of the segment, the average vector of all segments of the entire band is taken as the angle, and the dipole distance is taken as the length to obtain the segmented average vector area based on pole pitch.

[0095] The following method can also be used to calculate the segmented average vector area based on the pole distance: record the average of the main current direction of all frames in each segment as the average vector of the segment, take the average vector of all segments of the entire band as the angle, and take the main current intensity as the length to obtain the segmented average vector area based on the pole distance:

[0096] The calculation method of the segmented average vector area based on amplitude is as follows: the average of the dipole vector angles of all frames of each segment is recorded as the average vector of the segment, the average vector of all segments of the entire band is taken as the angle, and the magnetic field amplitude is taken as the length to obtain the segmented average vector area based on the pole pitch.

[0097] See also Figure 5 , Figure 6 , Figure 5 It is a schematic diagram of the segmented average vector area based on the pole distance. Figure 6 is a schematic diagram of the segmented average vector area based on amplitude. Figure 5 and Figure 6 As shown, due to Figure 5 and Figure 6 It is calculated based on the angle swept by the average vector, not the dipole vector angle in each frame, so there is no Figure 3 and Figure 4 In the jump. Figure 5 and Figure 6 The angle is the mean of the dipole vector angles, so Figure 5 and Figure 6 In the example, the origin of the coordinate axis is taken as the center, and the angle swept is the same; Figure 5 and Figure 6 The lengths used in Figure 5 and Figure 6The length corresponding to each angle is different.

[0098] The magnetic pole vector category feature includes at least one or more of segmented average quadrant distribution, segmented average quadrant level and segmented weighted quadrant distribution. The magnetic pole vector category feature is determined according to the average value of the target vector angle in each segment.

[0099] Specifically, the magnetic pole vector category feature is used to characterize the distribution characteristics of positive and negative magnetic pole vectors, and specifically includes one or more of the following three features:

[0100] Segment average quadrant distribution [nq1, nq2, nq3, nq4]: For the average dipole vector angle of each segment in the band, that is, the average vector in the above embodiment, the quadrant where each average vector is located is counted to obtain the number of average vectors in each quadrant, that is, the average quadrant distribution. Where nqi represents the number of average vectors in the i-th quadrant, i = 1, 2, 3, 4. For example, in the above Figure 3 In the figure, 2 average vectors fall in the first quadrant and 3 average vectors fall in the fourth quadrant, so the average quadrant distribution is [2,0,0,3].

[0101] Segmented average quadrant grade qs: Assign a corresponding grade to each quadrant, and obtain the segmented average quadrant grade according to the number of average vectors in each quadrant. For example, the first quadrant is assigned grade 1, the second quadrant is assigned grade 2, and so on. The segmented average quadrant grade qs is the type mark of each quadrant. For example, [2,0,0,4] and [3,0,0,1] both fall in quadrants 1 and 4, so their qs are both 14, and [0,1,1,0] falls in quadrants 2 and 3, so its qs is 23.

[0102] Segment-wise weighted quadrant distribution [vq1, vq2, vq3, vq4]: Weight the aforementioned segment-wise quadrant distribution according to the number of frames in each segment to obtain the segment-wise weighted quadrant distribution. Figure 3 In the example, there are 2 average vectors in the first quadrant, and the corresponding segmented frames of these two average vectors are z1 and z2 respectively. There are 3 average vectors in the fourth quadrant, and the corresponding segmented frames of these four average vectors are z3, z4 and z5 respectively. Then its average quadrant distribution is [2,0,0,3], and the segmented weighted quadrant distribution is [z1+z2,0,0,z3+z4+z5].

[0103] See also Figure 7 , Figure 7 A flowchart of a magnetocardiogram analysis method provided in another embodiment of the present application. Figure 7 As shown in FIG. 1 , in step 103, the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram is calculated based on multiple jump features, including:

[0104] S701, respectively calculating the fluctuation similarity, angle similarity, area similarity and magnetic pole vector similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the fluctuation category characteristics, angle category characteristics, area category characteristics and magnetic pole vector category characteristics.

[0105] S702 , performing weighted summation of the fluctuation similarity, the angle similarity, the area similarity and the magnetic pole vector similarity to obtain the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram.

[0106] In steps S701-S702, the similarity between the magnetocardiogram to be analyzed and the reference in each category is calculated according to the jump features of the four categories obtained in the above steps, and finally the similarity of each category is comprehensively analyzed to obtain the total similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram. There may be multiple references, and the magnetocardiogram to be analyzed is calculated for similarity with each target magnetocardiogram in turn, and then the heart point activity in the magnetocardiogram to be analyzed is comprehensively analyzed based on the calculation results to determine whether it is abnormal.

[0107] Specifically, the similarity calculation method for each category is as follows:

[0108]

[0109] Among them, p0 and n0 are basic parameters used to represent the characteristic parameters of healthy people, and a11, a12 and a13 are weight parameters.

[0110] It is understandable that if not all the fluctuation category features are obtained in the aforementioned steps, for example, only the jump wavelet range wg and the maximum clockwise jump angle pm are obtained, but the maximum counterclockwise jump angle nm is not obtained, then the above fluctuation similarity calculation formula can be adjusted, for example, the value of the weight parameter is adjusted, etc. The same is true for the following angle similarity, area similarity and magnetic pole vector similarity, which will not be repeated in the subsequent similarity calculation methods of each category.

[0111] Corner similarity rasi=a21×(r-r0)+a22×V+a23×Mse+a24×V1+a25×Mse1.

[0112] Among them, r0 is the basic parameter used to represent the characteristic parameters of healthy people, and a21 to a25 are weight parameters.

[0113] Area similarity arsi=a31×(ad-ar1)+a32×(af-ar2)+a33×(aread-ar3)+a34×(areaf-ar4).

[0114] Among them, ar1 to ar4 are basic parameters used to represent the characteristic parameters of healthy people, and a31 to a34 are weight parameters.

[0115] Magnetic pole vector similarity pvsi=a40×f41+b41×f42.

[0116] in,

[0117]

[0118] Among them, the quadrant level qs represents the different distribution types of the four quadrants, a41 to a44 and b41 to b44 are 0 or 1 according to the type of qs, and nqs, a40 and b40 are weight parameters.

[0119] The aforementioned basic coefficients and weight coefficients can be pre-set based on historical experience.

[0120] After obtaining the similarity of each category, it is summarized and analyzed. The specific calculation method is as follows:

[0121] Similarity cvsi=b0+b1×vosi+b2×rasi+b3×arsi+b4×pvsi.

[0122] Among them, coefficients b0 to b4 can be determined according to the target magnetocardiogram using a multiple linear regression method.

[0123] Optionally, the band includes at least one of the following: a QR band, an RS band, and a T band.

[0124] Specifically, the waveform of a cardiac cycle is usually composed of multiple bands, each of which corresponds to a different stage of cardiac electrical activity. Taking into account the characteristics of dipole jumps and the law of main current direction jumps in each band, bands or intervals with reasonable significance can be selected for the above-mentioned analysis steps. For example, for dipole jumps, at least one of the QR band, RS band and T band can be divided in the entire magnetocardiogram data set to be analyzed, and jump feature extraction and similarity comparison can be performed on the divided bands; and for the main current angle, since it has a similar jump phenomenon in the QR band, the QR band can be divided in the magnetocardiogram data set to be analyzed, and jump feature extraction and similarity comparison can be performed on the QR band.

[0125] See also Figure 8 , Figure 8 This is a flow chart of a magnetocardiogram analysis method provided by another embodiment of the present application. Figure 8 As shown in , the magnetocardiogram analysis method provided in the embodiment of the present application includes:

[0126] S801. Based on a magnetocardiogram dataset collected by a multi-channel magnetocardiograph, a magnetocardiogram dataset of a QR band, an RS band, and a T band is obtained.

[0127] S802 , based on the magnetocardiogram data sets of the QR band, RS band and T band, draw the magnetic field angle change curves and rotation angle change curves of the QR band, RS band and T band.

[0128] S803. Calculate the jump wavelet range wg of the QR band, RS band and T band based on the magnetic field angle change curve and the rotation angle change curve.

[0129] S804. Extract a series of dipole jump characteristic parameters based on the jump wavelet range wg of the QR band, RS band and T band.

[0130] S805: Determine the similarity between the magnetocardiogram image and the reference set based on a series of dipole jump characteristic parameters.

[0131] In this embodiment, firstly, a magnetocardiogram data set collected by a multi-channel magnetocardiograph is obtained, that is, a magnetocardiogram data set to be analyzed, and a magnetocardiogram data set of a QR band, an RS band, and a T band is respectively divided in the magnetocardiogram data set, and then the magnetic field angle change curve and the rotation angle change curve of each band are obtained, and the jump wavelet range wg of each band is calculated, and multiple different categories of dipole jump features are extracted based on the jump wavelet range. The specific calculation method is the same as the above steps and will not be repeated here. Thereafter, the similarity between the magnetocardiogram image and the reference image set (that is, the set of target magnetocardiograms) is determined based on all the dipole jump features of the QR band, the RS band, and the T band.

[0132] In addition, before step S801, a mean square error threshold value needs to be set in advance. For example, first, a data set of 1000 cases is constructed and divided into two groups, each with a 0 label and a 1 label, wherein the 0 label indicates that the data set corresponds to a healthy heart, and the 1 label indicates that the data set corresponds to a myocardial ischemic heart. The magnetocardiogram data of each data set in the QR band, RS band and T band are obtained respectively, and the magnetic field angle change curve and the rotation angle change curve are plotted. Taking the QR band as an example, 1000 magnetic field angle change curves and 1000 rotation angle change curves are plotted, and then the curves of the diseased data set and the curves of the healthy data set are analyzed, and according to the rules of the two, the mean square error threshold value m0 of the jump wavelet range wg of the QR band based on myocardial ischemia is set.

[0133] The magnetocardiogram analysis method provided in the embodiment of the present application sets four different categories of jump features on the basis of the first embodiment, and analyzes the magnetocardiogram data from four different directions. Compared with the magnetocardiogram analysis method in the prior art, the method of the present application quantitatively analyzes the magnetocardiogram data from different directions, and the results obtained are more comprehensive and more accurate. In addition, this embodiment specifically designs corresponding jump wavelet range determination methods and jump characteristic value calculation methods for the two vectors with jump conditions, namely the dipole vector and the main current. In actual application, a more suitable method can be adopted according to the specific application scenario and needs, thereby improving the breadth of applicability of the present application.

[0134] See also Fig. 9 , Fig. 9 This is a schematic diagram of the structure of a magnetocardiogram analysis device provided in an embodiment of the present application. Fig. 9 As shown in , the magnetocardiogram analysis device 900 includes:

[0135] The data acquisition module 910 is used to acquire the magnetocardiogram data set to be analyzed, and segment the magnetocardiogram data set into at least one band from the magnetocardiogram data set to be analyzed;

[0136] The turning angle set determination module 920 is used to determine a first turning angle set of the turning angle of the target vector corresponding to the band based on the magnetocardiogram data set of each band respectively; wherein the target vector is a vector formed by the magnetocardiogram dipoles in the isomagnetic map corresponding to the magnetocardiogram to be analyzed or a vector corresponding to the main current in the current density map corresponding to the magnetocardiogram to be analyzed; and the turning angle of the target vector is the turning angle of the target vector between two adjacent frames of isomagnetic maps or current density maps;

[0137] The jump wavelet range determination module 930 is used to determine the jump wavelet range of the band for each band of the magnetocardiogram data set based on the first rotation angle set of the rotation angle of the target vector corresponding to the band and the preset rotation angle fluctuation amplitude threshold; the jump wavelet range of the band is the rotation angle range of the target vector corresponding to the band in the continuous rotation process;

[0138] A jump feature extraction module 940 is used to determine, for each band of the magnetocardiogram data set, that the corners that meet the jump wavelet range are continuous corners and the other corners are jump corners, and determine a plurality of jump features according to the features of the continuous corners and the features of the jump corners;

[0139] The image analysis module 950 is used to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the multiple jump features, and obtain the analysis result of the magnetocardiogram to be analyzed according to the similarity.

[0140] Further, the plurality of jump features include at least one or more of a fluctuation category feature, an angle category feature, an area category feature, and a magnetic pole vector category feature;

[0141] Among them, the fluctuation category features are determined according to the continuous turning angles and jump turning angles; the turning angle category features are determined according to the continuous turning angles of each segment in the band, and the segments are divided according to the jump turning angles in the band; the area category features are determined according to the area swept by the continuous turning angles of each segment; the magnetic pole vector category features are determined according to the mean value of the target vector angle in each segment.

[0142] Further, the fluctuation category characteristics include at least one or more of the number of jumps, the jump wavelet range, the maximum clockwise jump angle, and the maximum counterclockwise jump angle;

[0143] The corner category features include at least one or more of the corner sum, angular velocity, corner mean square error, weighted angular velocity and weighted corner mean square error;

[0144] Area category features include piecewise vector area and / or piecewise average vector area;

[0145] The magnetic pole vector category features include at least one or more of a piecewise average quadrant distribution, a piecewise average quadrant rank, and a piecewise weighted quadrant distribution.

[0146] Furthermore, the image analysis module 950 is used to:

[0147] The fluctuation similarity vosi, the rotation angle similarity rasi, the area similarity arsi and the magnetic pole vector similarity pvsi are calculated according to the following formula:

[0148]

[0149] rasi=a21×(r-r0)+a22×V+a23×Mse+a24×V1+a25×Mse1;

[0150] arsi=a31×(ad-ar1)+a32×(af-ar2)+a33×(aread-ar3)+a34×(areaf-ar4);

[0151] pvsi=a40×f41+b40×f42;

[0152] in,

[0153]

[0154] Among them, p0, n0, a11, a12, a13, r0, a21, a22, a23, a24, a25, ar1, ar2, ar3, ar4, a31, a32, a33, a34, nqs, a40 and b40 are preset parameters, wg is the jump wavelet range, pm is the maximum clockwise jump angle, nm is the maximum counterclockwise jump angle, r is the rotation angle, V is the angular velocity, Mse is the rotation angle mean square error, V1 is the weighted angular velocity , Mse1 is the weighted angular mean square error, ad is the segmented vector area based on pole distance, af is the segmented vector area based on amplitude, areaad is the segmented average vector area based on pole distance, areaf is the segmented average vector area based on amplitude, nqi is the segmented average quadrant distribution of the i-th quadrant, vqi is the segmented weighted quadrant distribution of the i-th quadrant, i=1,2,3,4, a41 to a44 and b41 to b44 are 0 or 1 respectively according to the different distributions of the four quadrants;

[0155] The similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram is obtained by weighted summation of the fluctuation similarity, angle similarity, area similarity and magnetic pole vector similarity.

[0156] Further, when the turning angle set determination module 920 is used to determine the first turning angle set of the turning angle of the target vector corresponding to each band based on the magnetocardiogram data set of each band, the turning angle set determination module 920 is specifically used to:

[0157] Based on the magnetocardiogram data set of each band, the isomagnetic map of the band is drawn, and according to the magnetic field strength of each frame of the isomagnetic map, the vector angle of the magnetocardiogram dipole in each frame of the isomagnetic map is determined to obtain a first set of rotation angles of the magnetocardiogram dipole between every two frames of the isomagnetic map, and the first set is determined as the first rotation angle set of the rotation angle of the target vector corresponding to the band; or,

[0158] Based on the magnetocardiogram data set of each band, a current density map of the band is drawn, and according to the main current direction in each frame of the current density map, a second set of the main current angles between every two frames of the current density map is determined, and the second set is determined as the first angle set of the angle of the target vector corresponding to the band.

[0159] Further, the preset turning angle fluctuation amplitude threshold includes a mean square error threshold. When the jump wavelet range determination module 930 is used for determining the jump wavelet range of the band based on the set of turning angles of the target vector corresponding to the band and the preset turning angle fluctuation amplitude threshold for each band of the magnetocardiogram data set, the jump wavelet range determination module 930 is specifically used to:

[0160] Arrange the corners in the first corner set from large to small according to their absolute values ​​to obtain a second corner set;

[0161] Calculate the mean square errors of a preset number of adjacent corners in the second corner set respectively to obtain a mean square error set corresponding to the second corner set;

[0162] The mean square error in the mean square error set that is smaller than the preset angle fluctuation amplitude threshold is determined as the small fluctuation mean square error;

[0163] The corners in the second corner set corresponding to the small fluctuation mean square error are determined as the small fluctuation corners in the second corner set, and the jump wavelet range is determined according to the maximum value of the small fluctuation angles.

[0164] Further, the band includes at least one of the following: a QR band, an RS band, and a T band.

[0165] It should be noted that for other corresponding descriptions of the functional modules involved in the magnetocardiogram analysis device provided in the embodiment of the present application, reference can be made to the corresponding descriptions in the above method, which will not be repeated here.

[0166] See also Fig.10 , Fig.10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Fig.10 As shown in , the electronic device 1000 includes a processor 1010 , a memory 1020 , and a bus 1030 .

[0167] The memory 1020 stores machine-readable instructions executable by the processor 1010. When the electronic device 1000 is running, the processor 1010 communicates with the memory 1020 via the bus 1030. When the machine-readable instructions are executed by the processor 1010, the steps of the magnetocardiogram analysis method described above can be executed. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0168] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the magnetocardiogram analysis method described above can be executed. The specific implementation method can be found in the method embodiment, which will not be repeated here.

[0169] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0170] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0171] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0172] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0173] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0174] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes 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. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A magnetocardiogram analysis method, characterized in that: The method comprises: Acquire a magnetocardiogram data set to be analyzed, and segment the magnetocardiogram data set into at least one band from the magnetocardiogram data set to be analyzed; Based on the magnetocardiogram data set of each band, a first rotation angle set of the rotation angle of the target vector corresponding to the band is determined; wherein the target vector is a vector formed by the magnetocardiogram dipoles in the isomagnetic map corresponding to the magnetocardiogram to be analyzed or a vector corresponding to the main current in the current density map corresponding to the magnetocardiogram to be analyzed; the rotation angle of the target vector is the rotation angle of the target vector between two adjacent frames of isomagnetic maps or current density maps; For each band of the magnetocardiogram data set, based on a first rotation angle set of a target vector corresponding to the band and a preset rotation angle fluctuation amplitude threshold, a jump wavelet range of the band is determined; the jump wavelet range of the band is an angle range of the target vector corresponding to the band in a continuous rotation process; For each band of the magnetocardiogram data set, the angles satisfying the jump wavelet range of the band are determined as continuous angles, and the other angles are jump angles, and a plurality of jump features are determined according to the characteristics of the continuous angles and the characteristics of the jump angles; the plurality of jump features include fluctuation category features, angle category features, area category features, and magnetic pole vector category features; Calculating the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the plurality of jump features, and obtaining an analysis result of the magnetocardiogram to be analyzed according to the similarity; The calculating the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the multiple jump features includes: The fluctuation similarity vosi, the rotation angle similarity rasi, the area similarity arsi and the magnetic pole vector similarity pvsi are calculated according to the following formula: rasi=a21×(r-r0)+a22×V+a23×Mse+a24×V1+a25× Mse1; arsi=a31×(ad-ar1)+a32×(af-ar2)+a33×(aread-ar3)+a34×(areaf-ar4); pvsi=a40×f41+b40×f42; in, Among them, p0, n0, a11, a12, a13, r0, a21, a22, a23, a24, a25, ar1, ar2, ar3, ar4, a31, a32, a33, a34, nqs, a40 and b40 are preset parameters, wg is the jump wavelet range, pm is the maximum clockwise jump angle, nm is the maximum counterclockwise jump angle, r is the angle, V is the angular velocity, Mse is the mean square error of the angle, V1 is the weighted angular velocity, Mse1 is the weighted angular velocity. is the weighted angular mean square error, ad is the segmented vector area based on the pole distance, af is the segmented vector area based on the amplitude, areaad is the segmented average vector area based on the pole distance, areaf is the segmented average vector area based on the amplitude, nqi is the segmented average quadrant distribution of the i-th quadrant, vqi is the segmented weighted quadrant distribution of the i-th quadrant, i=1,2,3,4, a41 to a44 and b41 to b44 are 0 or 1 respectively according to the different distributions of the four quadrants; The fluctuation similarity, the rotation angle similarity, the area similarity and the magnetic pole vector similarity are weightedly summed to obtain the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram.

2. The magnetocardiogram analysis method according to claim 1, characterized in that: The fluctuation category feature is determined based on the continuous turning angle and the jump angle; the angle category feature is determined based on the continuous turning angle of each segment in the band, and the segments are divided according to the jump angle in the band; the area category feature is determined based on the area swept by the continuous turning angle of each segment; the magnetic pole vector category feature is determined based on the average value of the target vector angle in each segment.

3. The magnetocardiogram analysis method according to claim 2, characterized in that: The fluctuation category characteristics include at least one or more of the number of jumps, the range of the jump wavelet, the maximum clockwise jump angle, and the maximum counterclockwise jump angle; The corner category features include at least one or more of the corner sum, angular velocity, corner mean square error, weighted angular velocity and weighted corner mean square error; The area category features include segment vector area and / or segment average vector area; The magnetic pole vector category features include at least one or more of a piecewise average quadrant distribution, a piecewise average quadrant level, and a piecewise weighted quadrant distribution.

4. The magnetocardiogram analysis method according to claim 1, characterized in that: The determining of a first set of rotation angles of a target vector corresponding to each band based on the magnetocardiogram data set of each band respectively comprises: Based on the magnetocardiogram data set of each band, the isomagnetic map of the band is drawn, and according to the magnetic field strength of each frame of the isomagnetic map, the vector angle of the magnetocardiogram dipole in each frame of the isomagnetic map is determined to obtain a first set of rotation angles of the magnetocardiogram dipole between every two frames of the isomagnetic map, and the first set is determined as a first rotation angle set of the rotation angle of the target vector corresponding to the band; or, Based on the magnetocardiogram data set of each band, a current density map of the band is drawn, and according to the main current direction in each frame of the current density map, a second set of turning angles of the main current between every two frames of the current density map is determined, and the second set is determined as the first turning angle set of the target vector corresponding to the band.

5. The magnetocardiogram analysis method according to claim 1, characterized in that: The preset turning angle fluctuation amplitude threshold includes a mean square error threshold. For each band of the magnetocardiogram data set, the set of turning angles of the target vector corresponding to the band and the preset turning angle fluctuation amplitude threshold are used to determine the jump wavelet range of the band, including: Arrange the corners in the first corner set from large to small according to their absolute values ​​to obtain a second corner set; Respectively calculating mean square errors of a preset number of adjacent corners in the second corner set to obtain a mean square error set corresponding to the second corner set; Determine the mean square error in the mean square error set that is smaller than the preset turning angle fluctuation amplitude threshold as a small fluctuation mean square error; The corner in the second corner set corresponding to the small fluctuation mean square error is determined as the small fluctuation corner in the second corner set, and the jump wavelet range is determined according to the maximum value of the small fluctuation angles.

6. The magnetocardiogram analysis method according to claim 1, characterized in that: The band includes at least one of the following: a QR band, an RS band, and a T band.

7. A magnetocardiogram analysis device, characterized in that: The magnetocardiogram analysis device comprises: A data acquisition module, used for acquiring a magnetocardiogram data set to be analyzed, and segmenting the magnetocardiogram data set into at least one band from the magnetocardiogram data set to be analyzed; The turning angle set determination module is used to determine the first turning angle set of the turning angle of the target vector corresponding to the band based on the magnetocardiogram data set of each band respectively; wherein the target vector is the vector formed by the magnetocardiogram dipoles in the isomagnetic map corresponding to the magnetocardiogram to be analyzed or the vector corresponding to the main current in the current density map corresponding to the magnetocardiogram to be analyzed; the turning angle of the target vector is the turning angle of the target vector between two adjacent frames of isomagnetic maps or current density maps; A jump wavelet range determination module is used to determine the jump wavelet range of each band of the magnetocardiogram data set based on a first rotation angle set of the rotation angle of the target vector corresponding to the band and a preset rotation angle fluctuation amplitude threshold; the jump wavelet range of the band is the rotation angle range of the target vector corresponding to the band in the continuous rotation process; A jump feature extraction module is used for determining, for each band of the magnetocardiogram data set, that the angles satisfying the jump wavelet range of the band are continuous angles, and other angles are jump angles, and determining a plurality of jump features according to the features of the continuous angles and the features of the jump angles; the plurality of jump features include fluctuation category features, angle category features, area category features, and magnetic pole vector category features; An image analysis module, configured to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the plurality of jump features, and obtain an analysis result of the magnetocardiogram to be analyzed according to the similarity; When the image analysis module is used to calculate the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram according to the multiple jump features, the image analysis module is specifically used to: The fluctuation similarity vosi, the rotation angle similarity rasi, the area similarity arsi and the magnetic pole vector similarity pvsi are calculated according to the following formula: rasi=a21×(r-r0)+a22×V+a23×Mse+a24×V1+a25× Mse1; arsi=a31×(ad-ar1)+a32×(af-ar2)+a33×(aread-ar3)+a34×(areaf-ar4); pvsi=a40×f41+b40×f42; in, Among them, p0, n0, a11, a12, a13, r0, a21, a22, a23, a24, a25, ar1, ar2, ar3, ar4, a31, a32, a33, a34, nqs, a40 and b40 are preset parameters, wg is the jump wavelet range, pm is the maximum clockwise jump angle, nm is the maximum counterclockwise jump angle, r is the angle, V is the angular velocity, Mse is the mean square error of the angle, V1 is the weighted angular velocity, Mse1 is the weighted angular velocity. is the weighted angular mean square error, ad is the segmented vector area based on the pole distance, af is the segmented vector area based on the amplitude, areaad is the segmented average vector area based on the pole distance, areaf is the segmented average vector area based on the amplitude, nqi is the segmented average quadrant distribution of the i-th quadrant, vqi is the segmented weighted quadrant distribution of the i-th quadrant, i=1,2,3,4, a41 to a44 and b41 to b44 are 0 or 1 respectively according to the different distributions of the four quadrants; The fluctuation similarity, the rotation angle similarity, the area similarity and the magnetic pole vector similarity are weightedly summed to obtain the similarity between the magnetocardiogram to be analyzed and the target magnetocardiogram.

8. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the magnetocardiogram analysis method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the magnetocardiogram analysis method according to any one of claims 1 to 6 are executed.

Images