CN115581459B - LevKov-based multichannel fixed frequency noise filtering method, computer equipment and program product - Google Patents

Abstract

The application discloses a LevKov-based multichannel fixed frequency noise filtering method, a computer device and a program product, wherein the LevKov-based multichannel fixed frequency noise filtering method comprises the following steps: obtaining a plurality of magnetocardiogram waveform data by using the multichannel leads; obtaining a peak point in a nonlinear region predicted in the magnetocardiogram waveform data; dividing the nonlinear region and the linear region, comprising: shifting the peak point by a specified threshold value along the front and back of the time axis to obtain two starting points, taking the connecting line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, taking the data points with the peak point on two sides of the time axis and the maximum distance from the reference line on the corresponding side as end points, wherein a nonlinear region is arranged between the two end points, and the rest is a linear region; and filtering fixed frequency noise from the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

Description

Multi-channel fixed frequency noise filtering method based on LevKov, computer equipment and program product

Technical Field

The present application relates to the field of biomedical signal processing, and in particular to a multi-channel fixed-frequency noise filtering method based on LevKov, a computer device and a program product.

Background Art

At present, there are many clinical detection methods for heart diseases, including invasive tests such as coronary angiography and intracardiac electrophysiology, and non-invasive tests such as electrocardiogram and coronary artery CT. However, these detection technologies have shortcomings, such as being traumatic, causing radiation damage to the human body, or being expensive, so their clinical application is subject to certain restrictions.

Commonly used electrocardiograms are non-invasive and simple, but they have defects such as low sensitivity and accuracy. Magnetocardiography (MCG), as a non-invasive and highly accurate detection technology, has become a research hotspot in recent years. Cardiac activity generates weak currents, which generate weak magnetic fields around them. Magnetocardiography is a detection method that collects and analyzes the changes in the surrounding magnetic field caused by cardiac excitement.

In 1963, Baule and McFee first detected the magnetic field signal generated by the heart using a coil magnetometer in an unshielded room at room temperature. In 1970, Cohen and others invented the superconducting quantum interference device (SQUID), which brought a revolutionary breakthrough to the development of MCG. The SQUID sensor uses low-temperature superconducting technology. At a low temperature of minus 269. C, the resistance of metal niobium becomes zero and becomes a superconductor. The superconductor is made into a superconducting ring and the corresponding parts of the superconducting ring are made into two extremely thin insulating layers (Josephson nodes). When the bias current passes through, the superconducting state is destroyed and the Josephson effect and signal are generated. After being taken out by the oscillator and amplified, obvious magnetic field information is obtained.

Magnetocardiographic instruments have made great progress in recent years. They have evolved from single-channel to multi-channel, from single-point measurement to multi-point measurement, and also have the function of detecting three-dimensional signals and realizing three-dimensional reconstruction of the heart. Magnetocardiographic instruments based on atomic magnetometers are not restricted by the conditions of working in low-temperature environments, have higher sensitivity, can be further miniaturized and cost-effective, and expand their scope of application.

However, the high sensitivity and high precision of the magnetocardiometer make it very susceptible to interference from external noise. In addition, the hospital environment is noisy, and various devices often generate certain fixed-frequency noise when turned on. When the fixed-frequency noise intersects with the effective frequency band of the magnetocardiogram signal, the traditional notch filter will filter out the part of the magnetocardiogram data that overlaps with the noise frequency band, causing the QRS wave to deform; the LevKov algorithm can effectively retain the signal of the part of the QRS wave that overlaps with the noise frequency band, but it is only used to remove high-frequency interference such as power frequency interference. When the signal-to-noise ratio is particularly poor, it is difficult to determine the noise template range and the template area cannot be accurately obtained; when the frequency of noise removal is low, it does not meet the assumption that its template area is linear, and the filtering effect is poor.

Summary of the invention

Based on this, it is necessary to provide a multi-channel fixed-frequency noise filtering method based on LEVKOV to address the above technical problems.

This application is based on LEVKOV's multi-channel fixed-frequency noise filtering method, including:

Using multi-channel leads to obtain multiple magnetocardiogram waveform data;

Obtaining a peak point in a predicted nonlinear region in the magnetocardiogram waveform data;

The nonlinear region and the linear region are divided, including: offsetting the peak point forward and backward along the time axis by a specified threshold value to obtain two starting points, using the line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, using the data points where the peak point is on both sides of the time axis and has the largest distance from the corresponding side reference line as endpoints, the area between the two end points is the nonlinear area, and the rest is the linear area;

Fixed frequency noise is filtered out in the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

Optionally, when the peak point is offset forward and backward along the time axis, it corresponds to a first specified threshold and a second specified threshold respectively, the magnetocardiogram waveform data includes a continuous heartbeat signal, and the sum of the first specified threshold and the second specified threshold is greater than the width of the QRS wave in the heartbeat signal.

Optionally, obtaining the peak point of the magnetocardiogram waveform data specifically includes:

Obtaining a peak location threshold, wherein the peak location threshold is less than an average value of each peak in the magnetocardiogram waveform data;

The peak point is determined according to the relative relationship with the peak location threshold.

Optionally, the peak location threshold is calculated by the formula:

The obtained formula is that the central magnetic image waveform data includes multiple statistical cycles, SecMax is a sequence after the maximum values of all statistical cycles are sorted, a is a threshold generation coefficient less than 1, n1 is the starting data selected in the sequence SecMax, and n2 is the ending data selected in SecMax.

Optional, fixed frequency noise filtering for linear region, including:

For one of the data points to be processed, the data point is filtered using the buffered data within a predetermined time; the buffered data is obtained from the magnetocardiogram waveform data to be processed and has a length of M, where M=FS/FRE _noise , where FS is the sampling rate and FRE _noise is the frequency of the fixed frequency noise;

The data point to be processed is the data point corresponding to the next moment of the cached data;

After filtering the data point, the first data data _filter is obtained:

Cache _n is the cache data;

datum is the data point to be processed.

Optionally, it also includes using the data at the median position of the cached data to make a difference with the first data to obtain noise data, and the set of the noise data is a noise template;

The cache data and the noise template are updated in real time:

The real-time updating of the cached data is to use the data point corresponding to the next moment of the cached data to be placed into the cached data, and remove the earliest data point on the time axis;

The real-time updating of the noise template is to place the latest noise data into the noise template and remove the earliest data point on the time axis.

Optionally, when fixed frequency noise is filtered out in a nonlinear region, the following steps are specifically included:

Subtracting the data point at the median position of the cached data from the oldest noise data in the noise template to obtain current filtered data;

The cache data and the noise template are updated in real time:

The real-time updating of the cached data is to use the data point corresponding to the next moment of the cached data to be placed into the cached data, and remove the earliest data point on the time axis;

The real-time updating of the noise template is to put the oldest noise data into the noise template and remove the oldest noise data.

Optionally, when fixed-frequency noise is filtered out for a nonlinear region, a linear region with a data length of M adjacent to the nonlinear region is filtered in accordance with the filtering method of the nonlinear region.

The present application also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the LEVKOV multi-channel fixed frequency noise filtering method described in the present application.

The present application also provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the LEVKOV multi-channel fixed frequency noise filtering method described in the present application.

The multi-channel fixed frequency noise filtering method based on LEVKOV in this application has at least the following effects:

The peak points obtained in the magnetocardiogram waveform data of the present application are used to divide the nonlinear area and the linear area. Since the peak points used for division come from the magnetocardiogram waveform data, the present application can adaptively and specifically complete the division of the obtained magnetocardiogram waveform data according to different noise environments.

The present application divides the nonlinear region and the linear region by confirming the two endpoints of the nonlinear region. On this basis, based on the division result, the linear region and the nonlinear region of each magnetocardiogram waveform data are respectively subjected to fixed frequency noise filtering, thereby improving the filtering effect.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic flow chart of a multi-channel fixed frequency noise filtering method based on LEVKOV in one embodiment of the present application;

FIG. 2 is a schematic diagram of the process of sub-steps of step S200 in FIG. 1 ;

FIG3 is a schematic diagram of the process of sub-steps of step S400 in FIG1 ;

FIG4 is a schematic diagram of the process of sub-steps of step S400 in FIG1 ;

FIG5 is a flowchart of a multi-channel fixed frequency noise filtering method based on LEVKOV in one embodiment of the present application;

FIG6 is a diagram showing the internal structure of a computer device in one embodiment;

FIG. 7 is original time domain data of magnetocardiogram provided in an embodiment of the present application;

FIG8 is the time domain data after superimposing 25 Hz noise in FIG7;

FIG9 is the time domain data of FIG7 after filtering by Chebyshev 25Hz band-stop filter;

FIG10 is the time domain data of FIG7 after filtering by an integral coefficient 25 Hz notch filter;

FIG11 is the time domain data after filtering of FIG7 using the multi-channel fixed frequency noise filtering method based on LEVKOV of the present application;

FIG12 is the original frequency domain data of the magnetocardiogram provided in one embodiment of the present application;

FIG13 is the frequency domain data of FIG12 after superimposing 25 Hz noise;

FIG14 is the frequency domain data of FIG12 after filtering by Chebyshev 25 Hz band-stop filter;

FIG15 is the frequency domain data of FIG12 after filtering by an integral coefficient 25 Hz notch filter;

FIG. 16 is the frequency domain data after filtering in FIG. 12 using the multi-channel fixed frequency noise filtering method based on LEVKOV of the present application.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

1 to 4 , a multi-channel fixed-frequency noise filtering method based on LevKov is provided in one embodiment of the present application, including steps S100 to S400 described below, and the sub-steps in each step are optional solutions as refinements and extensions of the technical solution.

Step S100, obtaining a plurality of magnetocardiogram waveform data using multi-channel leads;

In step S100, the multi-channel leads may be, for example, 36 leads. Step S100 includes data preprocessing (step S110) and lead selection (step S120).

In step S110, data preprocessing includes differential processing and integral processing performed in sequence, wherein:

The differential processing includes: performing differential processing at intervals of N. The specific formula is as follows:

X _dev (i)＝X(i+N)-X(i) Formula (3.1)

In the above formula, X _dev is the data after differential processing, X is the original magnetocardiographic data, and N is the differential interval coefficient. For example, when the sampling rate is 1000 Hz, N is 5-15, such as 10.

The integral processing includes: taking the absolute value of the difference and performing m-point window integration. The specific formula is as follows:

In the above formula, _Xwin is the final data after preprocessing, _Xdev is the data after differential processing, and m is the window integration length. For example, when the sampling rate is 1000hz, m is 140-150, such as 145.

In step S120, lead selection includes:

When the data is initialized, an integer number of leads in the central area of 20% to 40% of the multi-channel leads is selected as the noise template range positioning leads.

And/or selecting an integer number of leads with the highest 20% to 40% R wave amplitude in the heart beat template as leads for locating the noise template range.

The duration of data initialization as cached data includes at least one heartbeat cycle, for example, the duration of cached data is 4 seconds. The lead selection may be, for example, selecting 8 positioning leads from 36 leads.

Step S200, obtaining a peak point in a predicted nonlinear region in the magnetocardiogram waveform data;

In step S200, the peak point of the magnetocardiogram waveform data is obtained, which specifically includes step S210 and step S220.

Step S210, obtaining a peak location threshold, where the peak location threshold is less than an average value of each peak in the magnetocardiogram waveform data.

Furthermore, the peak location threshold can be, for example, calculated by formula (3.3):

The central magnetic waveform data in the formula includes multiple statistical cycles, SecMax is the sequence after the maximum values of all statistical cycles are sorted, a is the threshold generation coefficient less than 1, n1 is the starting data selected in the sequence SecMax, and n2 is the end data selected in SecMax. Preferably, the interference of data anomalies is eliminated by selecting the data in the middle of the sequence, such as the SecMax sequence has a total of 20 maximum value sorted data, n1 is the data ranked 5th, n2 is the data ranked 16th, and a total of 12.

Specifically, the maximum value of each statistical period of the positioning lead is obtained, and after caching for a certain period of time (for example, it can be 20 seconds), the cached data is sorted (the data size is sorted and screened), and the maximum value of the data in a predetermined time period (for example, it can be 12 seconds) is taken as the average and then multiplied by a coefficient a (a can be, for example, 0.5 to 0.8, such as 0.7) to obtain the peak positioning threshold thr.

Step S220, determining the peak point according to the relative relationship with the peak location threshold.

Since there may be noise interference in the magnetocardiogram waveform data, after obtaining the peak location threshold in this step, any area greater than the peak location threshold is identified as belonging to the nonlinear area, and the peak of the area is the determined peak point.

Step S300, dividing the nonlinear region and the linear region, including: offsetting the peak point by a specified threshold value along the time axis to obtain two starting points, using the line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, using the data points where the peak point is on both sides of the time axis and has the largest distance from the corresponding side reference line as endpoints, the area between the two end points is the nonlinear region, and the rest is the linear region;

Furthermore, each starting point is a data zero point on the magnetocardiogram waveform data where the time axis is unchanged and the data value is set to zero. The magnetocardiogram waveform data is represented by an x-y coordinate system, where the time axis is the x axis and the data value is the y axis. The data zero point refers to setting the data value to zero when the x-axis position remains unchanged. In this way, possible spikes near the nonlinear region are avoided, which may affect the confirmation of the left and right endpoints of the nonlinear region.

In step S300, when the peak point is shifted forward and backward along the time axis, the first specified threshold and the second specified threshold are respectively corresponding, the magnetocardiogram waveform data includes a continuous heartbeat signal, and the sum of the first specified threshold and the second specified threshold is greater than the width of the QRS wave (the nonlinear region where R is located) in the heartbeat signal. For example, the first specified threshold and the second specified threshold are both 80ms.

In step S300, the nonlinear region and the linear region are obtained, specifically including: obtaining the endpoints of the nonlinear region using a local distance method, the endpoints including a left endpoint and a right endpoint;

The local distance method includes formula (3.4):

distance＝(y _start -y _end )*x+(x _end -x _start )*y+x _start *y _end -x _end *y _start , wherein (x, y) are the search coordinates of the magnetocardiogram waveform data, x is the time axis of the magnetocardiogram waveform data, and y is the data value of the magnetocardiogram waveform data; (x _start , y _start ) is the starting point of the search range, (x _end , y _end ) is the end point of the search range, and distance is the local distance from (x, y) to the line connecting the two points (x _start , y _start ) and (x _end , y _end );

Obtaining the left endpoint of the nonlinear region includes: setting the data zero point of the first specified threshold before the wave peak point time axis as the starting point of the search range, setting the wave peak point as the end point of the search range, and when the local distance is the maximum, the search coordinate is the left endpoint of the nonlinear region;

Obtaining the right endpoint of the nonlinear region includes: setting the peak point as the starting point of the search range, setting the data zero point of the second specified threshold after the peak point time axis as the end point of the search range, and when the local distance is the maximum, the search coordinate is the right endpoint of the nonlinear region. Obtaining the left and right endpoints of the nonlinear region means obtaining the nonlinear region, and the rest of the magnetocardiogram waveform data except the linear region is the linear region.

Furthermore, step S300 also includes that, in more than 50% of the positioning leads, the nonlinear regions of the leads are the same, and the nonlinear regions are considered to be obtained. Specifically, if more than 4 of the 8 positioning leads are determined to be non-template regions (non-linear regions), they are positioned as the final non-template regions, and the remaining data are noise template extraction regions (linear regions).

Step S400 , filtering out fixed frequency noise in the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

Step S400 includes steps S410 to S440.

Step S410, when filtering fixed frequency noise in the linear region, includes:

Step S411, for one of the data points to be processed, the data point is filtered using the buffered data within a predetermined time; the buffered data is obtained from the magnetocardiogram waveform data to be processed and has a length of M, M=FS/FRE _noise , where FS is the sampling rate, FRE _noise is the frequency of fixed frequency noise, and the data point to be processed is the data point corresponding to the next moment after the buffered data;

Step S412, filtering the data point to obtain first data data _filter , formula (3.5):

Cache _n is the cached data, and datum is the data point to be processed.

In step S410, filtering is performed for each data point, which is specifically implemented by the formula M=FS/FRE _noise . "Data points to be processed" refer to data points that are about to be cached but have not yet been cached.

It can be understood that when fixed-frequency noise is filtered out in the linear region, the cache data with an initial cache length of M is not filtered when the linear region is first entered. This is also reflected in formula (3.5), which means that the cached data is used to obtain the data after the fixed-frequency noise is filtered out.

Step S420 also includes using the data at the median position of the cached data to make a difference with the first data to obtain noise data. The set of noise data is a noise template. The cached data and the noise template are both updated in real time.

The noise data is obtained by the following formula:

temp＝Cache _n/2 -data _filter formula (3.6)

Wherein, Cache _n/2 is the data at the median position of the cache data, data _filter is the data after the fixed frequency noise is filtered out (first data), and temp is the obtained noise data.

The cache data and the noise template are updated in real time, specifically including step S421 and step S422.

Step S421, real-time updating of cached data is to use the data point corresponding to the next moment of the cached data to be placed into the cached data, and remove the earliest data point on the time axis (i.e., remove the oldest data point);

Step S422 , real-time updating of the noise template is to place the latest noise data into the noise template and remove the earliest data point on the time axis (ie, remove the oldest data point).

In step S420, the noise data is point data for the data point, and the noise template is a set of a plurality of noise data with a buffer length of M.

Step S430, when performing fixed frequency noise filtering for the nonlinear region, specifically includes: subtracting the data point at the median position of the cached data from the oldest noise data in the noise template to obtain the current filtered data, and both the cached data and the noise template are updated in real time.

The cache data and the noise template are updated in real time, specifically including step S431 and step S432. Among them:

Step S431, real-time updating of cached data is to use the data point corresponding to the next moment of the cached data to be placed into the cached data, and remove the earliest data point (oldest data point) on the time axis;

Step S432 : Real-time updating of the noise template is to place the oldest noise data into the noise template and remove the oldest noise data.

When fixed frequency noise is filtered out in the nonlinear region, the real-time update of the cache data is the same as that in the linear region, but the real-time update of the noise template is different. In the nonlinear region, the formula (3.5) for obtaining the first data data _filter is no longer applicable, and new noise data cannot be obtained through formula (3.6). At this time, the noise data in the noise template is recycled.

In step S432, filtering is completed by cyclically using the noise data in the noise template. For example, the noise template includes four noise data abcd, where a is the oldest noise data. a is placed in the noise template, and the oldest noise data a is removed. The four noise data in the updated noise template are bcda, and b is the oldest noise data.

In step S440, when fixed frequency noise is filtered out for the nonlinear region, a linear region with a data length of M adjacent to the nonlinear region is filtered according to the filtering method of the nonlinear region.

It can be understood that step S430 and step S440 cooperate with each other to keep the magnetocardiogram waveform data continuous after the fixed frequency noise is filtered out.

In each embodiment of the present application, by taking advantage of the magnetocardiogram multi-channel, the channel with the highest signal-to-noise ratio is selected as the positioning lead to determine the linear region and the nonlinear region (step S120). By means of the heartbeat signal in the magnetocardiogram waveform data, the peak positioning threshold thr is dynamically generated to preliminarily obtain the location of the nonlinear region (steps S210 and S220). Combined with the local distance method, the range of the nonlinear region is accurately located (step S300).

In each embodiment of the present application, the linear region and the nonlinear region are filtered respectively, and a good filtering effect can be achieved while retaining the QRS wave signal (step S400). Among them, when the fixed frequency noise is filtered out for the linear region, a noise template is also obtained (step S420 to step S422). The noise template is used to remove the noise in the nonlinear region (step S430 to step S432), ensuring that while removing the noise, the valid signal overlapping with the noise frequency band is retained.

It should be understood that, although the steps in the flowcharts of Figures 1 to 4 are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in Figures 1 to 4 may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

Referring to FIG5 , the flowchart of the multi-channel fixed frequency noise filtering method based on LEVKOV in each embodiment of the present application is briefly described as follows: after obtaining the data (i.e., magnetocardiogram waveform data) (step S100): on the one hand, data preprocessing (step S110), positioning lead selection (step S120), and template range positioning (step S300 dividing the nonlinear region and the linear region) are sequentially performed. After completion, linear region filtering (i.e., fixed frequency noise filtering) and template extraction (i.e., noise template extraction) are performed on each magnetocardiogram waveform data one by one, as well as non-template region (i.e., nonlinear region) filtering (step S400).

Referring to FIG6 , in one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG6 . The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. 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 and 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 operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store at least one of the following four: magnetocardiogram waveform data before filtering, cache data, noise data, and magnetocardiogram waveform data after filtering. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a multi-channel fixed-frequency noise filtering method based on LEVKOV is implemented.

In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

Step S100, obtaining a plurality of magnetocardiogram waveform data using multi-channel leads;

Step S200, obtaining a peak point in a predicted nonlinear region in the magnetocardiogram waveform data;

Step S300, dividing the nonlinear region and the linear region, including: offsetting the peak point by a specified threshold value along the time axis to obtain two starting points, using the line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, using the data points where the peak point is on both sides of the time axis and has the largest distance from the corresponding side reference line as endpoints, the area between the two end points is the nonlinear region, and the rest is the linear region;

Step S400 , filtering out fixed frequency noise in the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

Step S100, obtaining a plurality of magnetocardiogram waveform data using multi-channel leads;

Step S200, obtaining a peak point in a predicted nonlinear region in the magnetocardiogram waveform data;

Step S300, dividing the nonlinear region and the linear region, including: offsetting the peak point by a specified threshold value along the time axis to obtain two starting points, using the line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, using the data points where the peak point is on both sides of the time axis and has the largest distance from the corresponding side reference line as endpoints, the area between the two end points is the nonlinear region, and the rest is the linear region;

Step S400 , filtering out fixed frequency noise in the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

In one embodiment, a computer program product is provided, comprising computer instructions, which when executed by a processor implement the following steps:

Step S100, obtaining a plurality of magnetocardiogram waveform data using multi-channel leads;

Step S200, obtaining a peak point in a predicted nonlinear region in the magnetocardiogram waveform data;

Step S300, dividing the nonlinear region and the linear region, including: offsetting the peak point by a specified threshold value along the time axis to obtain two starting points, using the line between each starting point and the peak point as a reference line, traversing the data points between the peak point and each starting point, using the data points where the peak point is on both sides of the time axis and has the largest distance from the corresponding side reference line as endpoints, the area between the two end points is the nonlinear region, and the rest is the linear region;

Step S400 , filtering out fixed frequency noise in the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

In this embodiment, the computer program product includes a program code portion for executing the steps of the multi-channel fixed frequency noise filtering method based on LEVKOV in each embodiment of the present application when the computer program product is executed by one or more computing devices. The computer program product can be stored on a computer-readable recording medium. The computer program product can also be provided for download via a data network (e.g., via a RAN, via the Internet and/or via an RBS). Alternatively or additionally, the method can be encoded in a field programmable gate array (FPGA) and/or an application specific integrated circuit (ASIC), or the functionality can be provided for download with the aid of a hardware description language.

Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, 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 embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this 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 and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), 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).

In addition, the applicant has verified the feasibility and effectiveness of the technical solutions provided in each embodiment of the present application through experiments. Specifically, a 25Hz sine wave simulated noise is superimposed on the clean magnetic cardiology signal, and the Chebyshev 25Hz band-stop filter, the integer coefficient 25Hz notch filter and the LEVKOV-based multi-channel fixed frequency noise filtering method provided in each embodiment of the present application are used to compare the time domain and frequency domain results of the three.

Referring to Figures 7 to 16, it can be seen from the data in each figure that the Chebyshev 25HZ band-stop filter and the integer coefficient 25Hz notch filter will attenuate the signal in the part of the magnetocardiographic signal that overlaps with the interference frequency band, resulting in deformation of the magnetocardiographic waveform. The multi-channel fixed-frequency noise filtering method based on LEVKOV provided in each embodiment of the present application can effectively avoid deformation of the magnetocardiographic waveform and remove noise interference while retaining the effective signal.

The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. When the technical features in different embodiments are embodied in the same figure, it can be regarded that the figure also discloses the combination examples of the various embodiments involved.

The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims (7)

1. The LevKov-based multichannel fixed frequency noise filtering method is characterized by comprising the following steps of:

Obtaining a plurality of magnetocardiogram waveform data by using the multichannel leads;

Obtaining peak points in a nonlinear region predicted in the magnetocardiogram waveform data comprises the following steps: the peak location threshold is obtained by: the waveform data of the central magnetic diagram comprises a plurality of statistical periods, For sequences ordered by the maximum of all statistical periods,Generating coefficients for threshold values less than 1, n1 being the start data selected in the sequence SecMax and n2 being the end data selected in SecMax; all areas larger than the peak positioning threshold value are considered to belong to a nonlinear area; the peak positioning threshold value is smaller than the average value of each peak in the magnetocardiogram waveform data, and the peak point is determined according to the relative relation with the peak positioning threshold value;

Dividing the nonlinear region and the linear region, comprising: when the peak point is respectively shifted along the front and back of the time axis, a first designated threshold value and a second designated threshold value are respectively shifted to obtain two starting points, wherein the magnetocardiogram waveform data comprise continuous heart beat signals, and the sum of the first designated threshold value and the second designated threshold value is larger than the width of a QRS wave in the heart beat signals; traversing data points between the peak points and the starting points by taking a connecting line between each starting point and the peak point as a reference line, taking data points with the peak points on two sides of a time axis and the maximum distance from the corresponding side reference line as end points respectively, wherein a nonlinear region is arranged between the two end points, and the rest is a linear region;

and filtering fixed frequency noise from the linear region and the nonlinear region of each magnetocardiogram waveform data one by one until all magnetocardiogram waveform data are processed.

2. The LevKov-based multichannel fixed frequency noise filtering method according to claim 1, wherein when the fixed frequency noise filtering is performed for a linear region, the method includes:

For one data point to be processed, filtering the data point by utilizing cache data in a preset time; the buffer data is obtained from the magnetocardiogram waveform data to be processed, the length is M, m=fs/FRE _noise, wherein FS is the sampling rate, FRE _noise is the frequency of fixed frequency noise;

The data points to be processed are the data points corresponding to the time after the data are cached;

Filtering the data point to obtain first data ：

Caching data for the data;

Is the data point to be processed.

3. The LevKov-based multichannel fixed frequency noise filtering method according to claim 2, further comprising obtaining noise data by using the data of the median position of the buffered data and the first data as a difference value, wherein the set of noise data is a noise template;

the cache data and the noise template are updated in real time:

The real-time updating of the cache data is that the data point corresponding to the later time of the cache data is put into the cache data, and the earliest data point of a time axis is removed;

the real-time updating of the noise template is to put the latest noise data into the noise template and remove the earliest data point of the time axis.

4. The method for filtering multi-channel fixed frequency noise based on LevKov as set forth in claim 3, wherein the filtering of fixed frequency noise for a nonlinear region specifically includes:

subtracting the data point of the median position of the cache data from the oldest noise data in the noise template to obtain current filtering data;

the cache data and the noise template are updated in real time:

The real-time updating of the cache data is that the data point corresponding to the later time of the cache data is put into the cache data, and the earliest data point of a time axis is removed;

the real-time updating of the noise template is to put the oldest noise data into the noise template and remove the oldest noise data.

5. The method for filtering multi-channel fixed frequency noise based on LevKov as claimed in claim 4, wherein when the fixed frequency noise filtering is performed on a nonlinear region, filtering is performed on a linear region with a data length M adjacent to the nonlinear region according to a filtering mode of the nonlinear region.

6. Computer device comprising a memory, a processor and a computer program stored on the memory, characterized in that the processor executes the computer program to implement the steps of the LevKov-based multichannel fixed-frequency noise filtering method of any one of claims 1 to 5.

7. Computer program product comprising computer instructions which, when executed by a processor, implement the steps of the LevKov-based multichannel fixed-frequency noise filtering method according to any one of claims 1 to 5.