Physiological electrical signal downsampling method and evoked potential mapping machine

Abstract

The present application belongs to the technical field of physiological electrical signal detection, and particularly relates to a physiological electrical signal down-sampling method, which first divides the whole data of a physiological electrical signal into multiple buckets to obtain a set bucket, a front bucket and a rear bucket corresponding to the set data and the front data and the rear data of the set data; for the set bucket, a feature point is used as a sampling point of the set bucket; the first sampling point of the set bucket is used to iteratively obtain the sampling points of the front bucket in sequence; and the last sampling point of the set bucket is used to iteratively obtain the sampling points of the rear bucket in sequence. The waveform shape near the feature point is better retained, the amplitude and the latent period of the feature point after down-sampling are least affected, and the down-sampling distortion of a specific period can be effectively reduced for an induced electrical signal.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of physiological electrical signal detection, and particularly relates to a physiological electrical signal down-sampling method and an evoked potential graph machine. BACKGROUND

[0002] In the process of displaying evoked electrical waveforms of electroencephalogram and electromyogram, the abscissa often needs to be adjusted. In order to display original data of different lengths on a screen with fixed pixels, the sampling rate of the original data needs to be adjusted, that is, down-sampling processing. Decimation of the original data is a common down-sampling processing method. For example, patent No. CN 115114270B discloses a data down-sampling method and device, electronic equipment and computer readable medium, which performs down-sampling processing based on a maximum triangle three-bar decimation method (LTTB algorithm decimation). For each data segment, the data in each data segment is uniformly divided into multiple barrels according to a preset sampling ratio. The sampling representative points of each barrel are determined according to the data in each barrel. The sampling representative points of each barrel of each data segment form the down-sampling data corresponding to the original data. However, in the process of down-sampling of electroencephalogram and electromyogram, it is often necessary to mark feature points such as evoked potential peak values. The amplitude and latency of these feature points are used for disease diagnosis. The traditional LTTB algorithm decimation is a fixed interval decimation down-sampling method, which may cause the feature points not to be decimated, resulting in a change in the waveform amplitude after down-sampling, a large visual difference between the waveforms before and after processing, and an impact on the diagnosis accuracy. SUMMARY

[0003] The present application provides a physiological electrical signal down-sampling method and an evoked potential graph machine to reduce sampling distortion.

[0004] To solve the above technical problems, the present application provides a physiological electrical signal down-sampling method, which comprises: dividing the whole segment of physiological electrical signal data into multiple barrels to obtain the barrels corresponding to the set segment data and its preceding and subsequent segment data, i.e., the set barrel, the preceding barrel and the subsequent barrel; obtaining the sampling points of the set barrel; iteratively obtaining the sampling points of the preceding barrel from the first sampling point of the set barrel; and iteratively obtaining the sampling points of the subsequent barrel from the last sampling point of the set barrel.

[0005] Further, dividing the whole segment of data into multiple barrels comprises: according to the target sampling number setting the total number of barrels of the whole segment of data ; obtaining the number of barrels of the set barrel according to the feature points ; calculating the number of barrels of the preceding barrel ; calculating the number of barrels of the subsequent barrel ; and defining .

[0006] Further, the number of barrels of the set segment data is obtained according to the feature points This includes: marking feature points on a set segment of data; dividing the set segment of data corresponding to a single feature point into an independent set bucket, i.e., a feature bucket; dividing the set segment of data without a corresponding feature point into multiple set buckets, i.e., sampling buckets; obtaining sampling points for the set buckets includes: for feature buckets, using feature points as sampling points; for sampling buckets, using a traversal sampling method to obtain sampling points.

[0007] Further, calculate the number of buckets in the front bucket. Includes: obtaining and counting the information points of the front bucket. Starting from the first sampling point of the designated bucket, the information points of the preceding buckets are divided sequentially forward to obtain multiple preceding buckets. The width of the first preceding bucket is then determined. for The remainder, the width of the remaining front bucket for Rounding down.

[0008] Furthermore, obtaining the sampling points of the preceding bucket by iteratively moving forward from the first sampling point of the designated bucket includes: setting the sampling sequence number of the designated bucket corresponding to the first sampling point as (c+1), then the sampling sequence number of the last preceding bucket is... Determine if there are feature points in the last bucket. If so, use the feature points as sampling points in the last bucket. If not, use them as sampling buckets and use the traversal sampling method to obtain sampling points. Iterate forward in sequence until all sampling points in the last bucket are obtained.

[0009] Furthermore, calculate the number of buckets in the back bucket. Includes: obtaining and counting the information points of the back bucket. Starting from the last sampling point of the set bucket, the information points of the subsequent buckets are divided sequentially to obtain multiple subsequent buckets. The width of the last subsequent bucket is then determined. for The remainder, the width of the remaining bucket for Rounding down.

[0010] Furthermore, the process of iterating backward from the last sampling point of the set bucket to obtain the sampling points of the subsequent buckets includes: setting the sampling sequence number of the set bucket corresponding to the last sampling point as (c+b), then the sampling sequence number of the first subsequent bucket is (c+b+1); determining whether there is a feature point in the first subsequent bucket, if so, using the feature point as the sampling point of the first subsequent bucket; if not, using it as a value extraction bucket and obtaining the sampling point using the traversal value extraction method; iterating backward in sequence until the sampling points of all subsequent buckets are obtained.

[0011] Furthermore, the physiological electrical signal is configured as an evoked potential signal, and the set segment data is obtained based on the amplitude and latency; the feature points include at least one of marked points and outliers; wherein the marked points are configured to visually inspect and mark the start point, peak, trough, and end point of the feature waves in the set segment data; the outliers are configured to be obtained using an automatic identification method, including: calculating the mean of the baseline data. With variance Set the range of normal points as follows Points outside the specified range are considered outliers; and when there are multiple feature points in the current or subsequent bucket, the priority for selecting sampling points is that labeled points are greater than outliers.

[0012] In a second aspect, the present invention provides an induced potential mapping machine, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the method described above.

[0013] A display, connected to the processor, shows the waveform of the evoked potential signal after downsampling.

[0014] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method.

[0015] Fourthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method.

[0016] The beneficial effects of this invention are as follows: The method for downsampling physiological electrical signals first divides the entire segment of physiological electrical signal data into multiple buckets, obtaining a set segment of data and the buckets corresponding to its preceding and following segments, namely, a set bucket, a preceding bucket, and a following bucket; then, it obtains the sampling points of the set bucket; iterates forward using the first sampling point of the set bucket to obtain the sampling points of the preceding bucket; and iterates backward using the last sampling point of the set bucket to obtain the sampling points of the following bucket. In the set segment data of important time periods, feature points are used to optimize the sampling points, while keeping the total number of buckets unchanged. This not only ensures that the downsampling rate meets the requirements but also better preserves the waveform shape near the feature points. After downsampling, the amplitude and latency of the feature points are minimally affected, effectively reducing downsampling distortion in specific time periods for evoked signals.

[0017] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. To make the foregoing objects, features, and advantages of the invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 Yes, it's a flowchart of the physiological electrical signal downsampling method.

[0020] Figure 2 This is an example diagram of somatosensory evoked potential (SEP) acquisition and waveform.

[0021] Figure 3 This is a schematic diagram of the characteristic points of the SEP waveform.

[0022] Figure 4 This is a flowchart illustrating the iteration process of the defined segment.

[0023] Figure 5 This is a flowchart illustrating the process of the front bucket iteration.

[0024] Figure 6 This is a flowchart illustrating the iteration process of the back bucket.

[0025] Figure 7 This is a block diagram of the induced potential acquisition system.

[0026] Figure 8 This is a comparison chart showing the changes in peak / trough amplitude caused by traditional downsampling methods.

[0027] Figure 9 This is a comparison chart showing how current downsampling methods ensure constant amplitude of peaks / troughs.

[0028] Figure 10 This is a schematic diagram illustrating how the traditional downsampling method is affected by changes in the iteration start point.

[0029] Figure 11 This is a schematic diagram illustrating that the current downsampling method is unaffected by changes in the iteration start point.

[0030] Figure 12 This is a schematic diagram showing how traditional downsampling methods can disrupt the shape of characteristic waves.

[0031] Figure 13 This is a schematic diagram illustrating how current downsampling methods preserve the shape of characteristic waves. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] In related technologies, when the original signal consists of identical data spliced ​​together, the amplitudes (peak or valley values) of the feature points are the same. After downsampling using the LTTB algorithm, the amplitudes of the feature points may decrease to varying degrees, potentially leading to higher data distortion (e.g., ...). Figure 8 (As shown). Therefore, in at least one embodiment, a method for downsampling physiological electrical signals, namely a segmented downsampling method, is provided, comprising: dividing the entire segment of physiological electrical signal data into multiple buckets to obtain a set segment of data and the buckets corresponding to its preceding and following segments, namely, a set bucket, a preceding bucket, and a following bucket; obtaining the sampling points of the set bucket; using the first sampling point of the set bucket to iteratively obtain the sampling points of the preceding bucket; and using the last sampling point of the set bucket to iteratively obtain the sampling points of the following bucket. See Figures 1-6 Taking the evoked potential signal in physiological electrical signals as an example, the specific steps of downsampling are as follows.

[0034] Step S1: Divide the data into buckets.

[0035] For evoked electrical data, the amplitude and latency (time from the end of stimulation to the peak value) of the characteristic wave are used for disease diagnosis. A characteristic wave amplitude lower than normal or a latency that is too long indicates an abnormal peak value in the neural conduction pathway. For example... Figure 2 As shown, different characteristic waves exist in the signal when the stimulation site and signal acquisition site are different. For example, when stimulating the median nerve, the characteristic wave is N19 when acquiring data from the C3-Fz area of ​​the scalp. Therefore, the focus should be on the characteristic wave, and the consistency of the characteristic wave before and after downsampling should be prioritized. The data in this time period is the important time period data, i.e., the set segment data, and feature points are obtained from the set segment data. Of course, feature points can also be obtained from the entire data segment, but the workload will be much larger.

[0036] (1) Obtain feature points and information points.

[0037] For the defined data segment, the feature points include labeled points and outliers. The labeled points are configured to visually inspect and label the start point, peak, trough, and end point of characteristic waves in the defined data segment. The outliers are configured to be acquired using an automatic identification method, including: calculating the mean x̄ and variance δ of the baseline data; setting the range of normal points as [(x̄-3δ), (x̄+3δ)]; points outside this range are identified as outliers. Specifically, for example... Figure 3As shown, for data 0.1 seconds before stimulation, the mean x̄ and variance δ of the baseline data are calculated, and the range of normal points is set as [(x̄-3δ), (x̄+3δ)]. For non-defined data segments after stimulation, data exceeding the normal range are marked as abnormal data. Abnormal data consists of one or more continuous data segments, with the point in each segment deviating furthest from the normal range being the abnormal point. The information points can be obtained through an automatic identification method, specifically as follows: For the baseline, the baseline data 0.1 seconds before stimulation is acquired, and the mean x̄ and variance δ of the baseline data are calculated, with the range of the baseline set as [(x̄-δ), (x̄+δ)]. For peaks, the maximum peak value within the defined data segment is used; for troughs, the minimum trough value within the defined data segment is used. For the starting point of a characteristic wave, the point that enters the baseline range forward from the peak / trough with the shorter latency is used; for the ending point of a characteristic wave, the point that enters the baseline range backward from the peak / trough with the longer latency is used.

[0038] In some embodiments, the front-end data and back-end data can be obtained by setting segment data to acquire feature points and information points.

[0039] (2) Set the number of buckets.

[0040] Dividing the entire data segment into multiple buckets includes: based on the target number of samples. Set the total number of buckets for the entire data segment. ; Determine the number of buckets to be set based on feature points ; Calculate the number of buckets in the front bucket ; Calculate the number of buckets after the calculation ; and defined numerically .

[0041] (3) Divide and set buckets.

[0042] The number of buckets for a given segment of data is determined based on feature points. Includes: marking feature points on a defined segment of data and counting them. The data segment corresponding to a single feature point is divided into independent buckets, i.e., feature buckets, with the number of feature buckets being... That is, a single feature bucket contains only one feature point, but it may also contain additional information points or no information points; a bucket without feature points (i.e., containing only information points) The corresponding set segment data is divided into multiple set buckets, also known as extraction buckets, and the number of information points in each extraction bucket is not fixed. Definition The width of a bucket can be understood as the total number of information points and feature points within a single bucket. There are several ways to set the bucket width. For example, you can first set one feature point to correspond to one feature bucket, then the width of the feature bucket will be 1, meaning it contains only one feature point. Then, you can divide a set segment of data with no consecutive feature points into multiple sampling buckets, with the number of information points in each sampling bucket being... The rounding down is applied to cases where the width of the extraction bucket is not met, treating them as a single extraction bucket. For example, ... Figure 4 As shown, first, the width of the buckets is set in the defined data segment. Then, the number of feature points in each bucket is determined. If the number of feature points is zero, it is determined to be a value extraction bucket; if the number of feature points is 1, it is determined to be a feature bucket; if the number of feature points is greater than 1, it is determined to be a feature bucket. The feature buckets are then re-divided according to the feature points to ensure that each feature bucket contains only 1 feature point. The value extraction buckets are also re-divided to ensure that the total number of feature buckets and value extraction buckets equals the number of buckets in the defined data segment. .

[0043] (4) Divide the front bucket.

[0044] Calculate the number of buckets in front. Includes: obtaining and counting the information points of the front bucket. ; set the first sampling point of the bucket (e.g. Figure 5 Starting from the sampling point in the middle bucket (c+1), the information points of the front buckets are divided sequentially forward to obtain multiple front buckets. The width of the first or last front bucket is then... for The remainder, the width of the remaining front bucket for Rounding down to the nearest integer. For example... Figure 5 In the process, the width of bucket 1 (i.e. the first bucket) in the entire data segment is set to 1 information point, and the width of the remaining buckets is set to 3 information points.

[0045] (5) Divide the buckets.

[0046] Calculate the number of buckets after the calculation Includes: obtaining and counting the information points of the back bucket. ; to set the last sampling point of the bucket (e.g. Figure 5 Starting from the sampling point in the middle bucket (c+b), the information points of the subsequent buckets are divided sequentially to obtain multiple subsequent buckets. The width of the first or last subsequent bucket is then determined. for The remainder, the width of the remaining bucket for Rounding down to the nearest integer. For example... Figure 6 In the process, the width of bucket a (i.e., the last bucket) in the entire data segment is set to 1 information point, and the width of the remaining buckets is set to 3 information points.

[0047] In some embodiments, if the front-end data or the back-end data also contains feature points, the front bucket or the back bucket can be divided according to the set segment data division method.

[0048] Step S2: Obtain the sampling points of the designated bucket.

[0049] In the defined data segment, for the feature bucket, the feature points are directly used as sampling points; for the value extraction bucket, the sampling points are obtained by traversing the value extraction method.

[0050] For details, see Figure 4 The traversal sampling method includes: selecting the two buckets closest to the sampling bucket as constraint buckets; if the constraint bucket is a feature bucket, then the feature point is used as the constraint point; if the constraint bucket is a sampling bucket, then the average of all information points in the bucket is used as the constraint point; selecting any information point in the sampling bucket as the sampling point; forming a triangle with the sampling point and the two constraint points, and calculating the area of ​​the triangle; traversing all information points in the sampling bucket, and retaining the sampling point with the largest triangle area as the sampling point of the sampling bucket.

[0051] Step S3: Obtain the sampling points of the front bucket.

[0052] See Figure 5 The process of obtaining the sampling points of the preceding bucket by iterating backward from the first sampling point of the first bucket includes: setting the sampling sequence number of the bucket corresponding to the first sampling point as (c+1), then the sampling sequence number of the last preceding bucket is... ,Right now Figure 5 Buckets of mid-to-front-end data ; Determine if a feature point exists in the first bucket before the last bucket. If so, use the feature point as a sampling point for the first bucket before the last bucket; otherwise, use it as a sampling bucket and obtain sampling points using a traversal sampling method; iterate forward in this way until sampling points of all buckets are obtained. Specifically, as follows... Figure 5 In the process, starting from bucket c, the sampling points in bucket (c+1) and the mean value (i.e., the mean point) of the information points in bucket (c-1) are used as two constraint points for bucket c. Any information point in bucket c is selected as the sampling point. A triangle is formed by the sampling point and the two constraint points, and the area of ​​the triangle is calculated. All information points in the sampling bucket are traversed, and the sampling point with the largest triangle area is retained as the sampling point of bucket c. Then, the sampling points of each previous bucket are obtained sequentially along the iteration direction of the previous bucket, i.e., to the left.

[0053] Step S3: Obtain the sampling points of the back bucket.

[0054] See Figure 6 The process of obtaining sampling points for subsequent buckets by iterating backwards from the last sampling point of a given bucket includes: setting the sampling sequence number of the bucket corresponding to the last sampling point as (c+b), then the sampling sequence number of the first subsequent bucket is (c+b+1), which is bucket (c+b+1) in the subsequent data; determining whether there is a feature point in the first subsequent bucket; if so, using the feature point as the sampling point of the first subsequent bucket; otherwise, using it as a sampling bucket and obtaining sampling points using a traversal sampling method; iterating backwards to obtain sampling points for all subsequent buckets. Specifically, as shown... Figure 6In the process, starting from bucket (c+b+1), the sampling points in bucket (c+b) and the mean value (i.e., the mean point) of the information points in bucket (c+b+2) are used as two constraint points for bucket (c+b+1). Any information point in bucket (c+b+1) is selected as the sampling point. A triangle is formed by the sampling point and the two constraint points, and the area of ​​the triangle is calculated. All information points in the sampling bucket are traversed, and the sampling point with the largest triangle area is retained as the sampling point of bucket (c+b+1). Then, the sampling points of each subsequent bucket are obtained sequentially along the iteration direction of the subsequent buckets, i.e., moving to the right.

[0055] In some embodiments, for sampling buckets containing only information points, the traversal sampling method is applicable to obtaining sampling points from the set bucket, the front bucket, and the back bucket. For feature buckets containing feature points, the set bucket, the front bucket, and the back bucket preferentially select feature points as sampling points. If there are multiple feature points, the selection priority of sampling points is that of labeled points are greater than that of outliers.

[0056] In some embodiments, an evoked potential grapher is also provided, comprising: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the method as described; and a display connected to the processor to display the waveform of the evoked potential signal after downsampling.

[0057] For details, see Figure 7 The induced potential mapping machine includes structural components and hardware circuits. The operating platform of the method is a lower-level machine module, a higher-level machine module, and a display module. The signal input is provided by an electrode sensor, an amplifier, and an analog-to-digital conversion module, as detailed below.

[0058] (1) Electrode sensors to collect raw simulated EEG signals.

[0059] (2) Amplifier module (hardware analog circuit) amplifies the original analog EEG signal.

[0060] (3) Analog-to-digital conversion module (hardware) converts the amplified analog EEG signal into a digital signal at a fixed sampling rate.

[0061] (4) Lower-level module (MCU / ARM / DSP hardware + embedded software), storing / forwarding digital signals.

[0062] (5) Host computer module (PC hardware + PC software) processes digital signals, adjusts downsampling parameters and runs downsampling methods, and displays them to users in conjunction with the UI workflow.

[0063] (6) Display module: presents the UI interface, allows users to operate the UI, set detection requirements, and present the waveforms before and after downsampling.

[0064] The processor can be a central processing unit (CPU), an ASIC, or one or more integrated circuits configured to implement embodiments of the present invention. In specific implementations, if the memory and processor are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a PCI bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be categorized as an address bus, a data bus, a control bus, etc. If the memory and processor are integrated onto a single chip, they can communicate with each other through an internal interface.

[0065] In some embodiments, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the method. The storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0066] In some embodiments, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the method. The computer program may include program code, which includes computer operation instructions and may be stored in a computer-readable storage medium. Based on this understanding, when the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be implemented in the form of a software product or sold or used as an independent product, the computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the method described in the various embodiments of the present invention.

[0067] Experimental example.

[0068] The raw data of the evoked waveforms of EEG were downsampled using both the traditional downsampling method (i.e., LTTB algorithm sampling, referred to as the traditional method) and the segmented downsampling method of this case (referred to as the current method). The feature point amplitudes and waveforms of the downsampled data were compared, and the results are as follows.

[0069] (1) Stability of characteristic point amplitude.

[0070] contrast Figure 10 and Figure 11 It can be seen that the feature point peak appears around 80ms. Therefore, respectively at Figure 9, Figure 10 The waveform around 80ms (i.e., within the black box) is locally magnified, corresponding to... Figure 9 , Figure 10 A magnified view on the right. (This is based on observation.) Figure 9 , Figure 10 The peak values ​​of the feature points show that the LTTB algorithm extracts values ​​because fluctuations in the starting point cause fluctuations in the amplitude of the feature points; the segmented downsampling method can ensure the stability of the amplitude of the feature points.

[0071] (2) The magnitude of the feature point amplitude.

[0072] contrast Figure 8 and Figure 9 It can be seen that the feature point amplitudes are the right three, appearing around 160ms, 420ms, and 680ms respectively. LTTB algorithm sampling: The feature point amplitudes after downsampling are lower than the original data; using data before and after downsampling for diagnosis will lead to different conclusions, potentially causing misdiagnosis; Segmented downsampling method: The feature point amplitudes remain unchanged.

[0073] (3) Waveform.

[0074] contrast Figure 12 and Figure 13 It can be seen that the LTTB algorithm results in severe waveform loss, with large amplitude biphase waveforms that can provide visual reference, while downsampling transforms them into single-phase waveforms. The segmented downsampling method, on the other hand, produces characteristic peaks that are closer to the original waveform, are smoother, and retains the original biphase shape even for high-amplitude waveforms.

[0075] Based on the above-described preferred embodiments of the present invention, and through the above description, those skilled in the art can make various changes and modifications without departing from the technical concept of the present invention. That is, the technical scope of the present invention is not limited to the contents of the specification.

Claims

1. A method of downsampling a physiological electrical signal, the method comprising: The method comprises the following steps: dividing the whole data of the physiological electrical signal into a plurality of buckets to obtain a set of data and its corresponding bucket, i.e. a set of buckets, a front bucket and a rear bucket; obtaining the sampling points of the set of buckets; iteratively obtaining the sampling points of the front bucket using the first sampling point of the set of buckets; and iteratively obtaining the sampling points of the rear bucket using the last sampling point of the set of buckets; dividing the whole data into a plurality of buckets comprises: According to the feature point, a number of barrels of the barrel is acquired ; According to the feature point, the number of buckets for acquiring the segment data is set Comprising: marking the feature points on the set of data; dividing the set of data corresponding to a single feature point into an independent set of buckets, i.e. a feature bucket; dividing the set of data corresponding to no feature point into a plurality of sets of buckets, i.e. a value extraction bucket; obtaining the sampling points of the set of buckets comprises: for the feature bucket, taking the feature point as the sampling point; for the value extraction bucket, obtaining the sampling points using the traversal value extraction method.

2. The method of claim 1, wherein dividing the whole data into a plurality of buckets further comprises: According to the target sampling number Setting the total number of buckets for the entire data ; calculating the number of buckets of the front bucket ; calculating the number of buckets of the post-bucket ; and numerically defined .

3. The method of claim 2, wherein Counting the number of buckets of a front bucket Comprising: acquire information points of the front bucket and count as ; The information points of the front barrel are sequentially divided from the first sampling point of the set barrel as a starting point, to obtain a plurality of front barrels, and the width of the first front barrel or the last front barrel is is the remainder, and the width of the remaining front barrel is is rounded down.

4. The method of claim 1, wherein iteratively obtaining the sampling points of the front bucket using the first sampling point of the set of buckets comprises: The sampling sequence number of the set bucket corresponding to the first sampling point is set as (n-1) mod N , and the sampling sequence number of the last front bucket is (n-1) mod N . determining whether there is a feature point in the last front bucket, if yes, taking the feature point as the sampling point of the last front bucket; if no, taking it as a value extraction bucket and obtaining the sampling points using the traversal value extraction method; iteratively obtaining the sampling points of the front bucket until all the front buckets are obtained.

5. The method of claim 2, wherein calculating the number of buckets of the post-bucket comprising: acquire information points of the back bucket and count as ; The information points of the rear barrels are divided in turn from the end sampling point of the set barrel as the starting point, to obtain a plurality of rear barrels, and the width of the first rear barrel or the last rear barrel is the remainder, and the width of the remaining rear barrels is is rounded down.

6. The method of claim 1, wherein iteratively obtaining the sampling points of the rear bucket using the last sampling point of the set of buckets comprises: The sampling sequence number of the first post-barrel is ( +1) when the sampling sequence number of the last sampling point of the setting barrel is set as ( ). determining whether there is a feature point in the first rear bucket, if yes, taking the feature point as the sampling point of the first rear bucket; if no, taking it as a value extraction bucket and obtaining the sampling points using the traversal value extraction method; iteratively obtaining the sampling points of the rear bucket until all the rear buckets are obtained.

7. The method of claim 4 or 6, wherein the physiological electrical signal is configured as an evoked potential signal, and the set of data is obtained according to the amplitude and latency; the feature points comprise at least one of the labeled points and the abnormal points; wherein the labeled points are configured to visually check and label the starting point, peak, trough and ending point of the characteristic wave in the set of data; the abnormal points are configured to be obtained using an automatic identification method, comprising: calculating a mean value of the baseline data and a variance , setting a range of normal points to , judging as an abnormal point if out of the range; and when there are multiple feature points in the current bucket or the rear bucket, the selection priority of the sampling points is that the labeled points are greater than the abnormal points.

8. An evoked potential mapping machine characterized by comprising: The device comprises: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the method according to any one of claims 1-7; a display connected to the processor to display the waveform of the evoked potential signal after down-sampling.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method according to any one of claims 1-7.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method according to any one of claims 1-7.

Images