An interference wave labeling algorithm and brain fluctuation graph

By applying instantaneous outliers and single-frequency interference signal recognition and processing algorithms on the brain fluctuation graph, the interference signals in the EEG data are automatically identified and marked, which solves the problem of low manual operation efficiency in the existing technology and improves data acquisition Efficiency and signal quality.

CN118787370BActive Publication Date: 2024-11-19SHENZHEN CORNLEY HI TECHCO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411262210.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-11-19
Estimated Expiration
2044-09-10

Smart Images

  • Figure CN118787370B_ABST
    Figure CN118787370B_ABST
Patent Text Reader

Abstract

The present invention discloses an interference wave marking algorithm and a brain fluctuation graph, wherein the interference wave marking algorithm includes an instantaneous abnormal value recognition and processing algorithm and a single-frequency interference signal recognition and processing algorithm; the instantaneous abnormal value recognition and processing algorithm recognizes the interference signal caused by the instantaneous abnormality by the amplitude deviating from the baseline of the average value of the brain wave; the single-frequency interference signal recognition and processing algorithm recognizes the single-frequency interference signal by the position of the maximum value in the spectrum after fast Fourier transformation. The interference wave marking algorithm and the brain fluctuation graph provided by the present invention can automatically recognize and mark the interference wave, reduce the workload of the operator, and improve efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of brain fluctuation graphs, in particular to an interference wave marking algorithm and a brain fluctuation graph. Background Art

[0002] The EEG fluctuation graph is a special device used for EEG signal collection and analysis, and is often used in medical projects to examine brain function. The characteristic of the EEG fluctuation graph is that it requires long-term sampling, and the sampling time is at least 10 minutes. However, since the EEG signal itself is very weak, and the voltage peak is below 100uV, it is inevitable to introduce some interference signals during the EEG signal collection process. For example, the subject's body movement, the ringing signal of the mobile phone, the vibration caused by the startup of large equipment, etc., can all affect the collection of EEG signals. In existing products, interference signals are manually marked little by little, but the collected data is usually very large, and manual operation alone is very inefficient. Summary of the invention

[0003] In order to overcome the shortcomings of the existing technology, an interference wave marking algorithm and an electroencephalogram are proposed, which can automatically identify and mark interference waves, reduce the operator's workload and improve efficiency.

[0004] An interference wave marking algorithm, the interference wave marking algorithm is applied to a brain fluctuation graph, and includes a first sub-algorithm and / or a second sub-algorithm;

[0005] For the EEG data, use any one of the first sub-algorithm and the second sub-algorithm to process it, or first use the first sub-algorithm to process it, and then use the second sub-algorithm to process it, or first use the second sub-algorithm to process it, and then use the first sub-algorithm to process it;

[0006] The first sub-algorithm is an instantaneous outlier identification and processing algorithm, and comprises the steps of:

[0007] S11, dividing the EEG data into a plurality of preset time periods, each preset time period including N brain wave signals; in each preset time period, obtaining an average value of the brain waves in the current preset time period; wherein N is a natural number greater than 1;

[0008] S12, within the current preset time period, recording the time at which the first EEG signal greater than or equal to the average value of the EEG wave and exceeding the preset high threshold value occurs as the interference start time;

[0009] S13, starting from the interference start time, recording the interference end time as the time when the EEG signal is less than or equal to the preset low threshold for M consecutive times; wherein M is a natural number greater than 1;

[0010] S14, recording the time between the interference start time and the interference end time as the interference time period;

[0011] S15, combining the interference time periods in each preset time period to obtain a first interference time period;

[0012] The second sub-algorithm is a single-frequency interference signal identification and processing algorithm, and comprises the steps of:

[0013] S21, dividing the EEG data into a plurality of preset time periods, each preset time period including a plurality of lead data;

[0014] S22, performing a fast Fourier transform on each lead data to obtain a frequency spectrum of the lead data, wherein each frequency spectrum includes a maximum value;

[0015] S23, if the maximum values ​​in each spectrum are located at the same position, the time period at the position is the interference time period;

[0016] S23, combining interference time periods in the plurality of lead data to obtain a second interference time period.

[0017] A brain fluctuation graph, which adopts the interference wave labeling algorithm as mentioned above.

[0018] The interference wave marking algorithm and brain fluctuation graph provided by the present invention respectively use an instantaneous outlier value recognition and processing algorithm and a single-frequency interference signal recognition and processing algorithm to process the instantaneous outlier value interference signal and the single-frequency interference signal in the electroencephalogram data, and can automatically mark the time period in which the interference signal is located, thereby reducing the operator's workload and improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the interference of instantaneous abnormal values ​​of EEG signals;

[0020] Figure 2 Schematic diagram of single-frequency interference signal of EEG;

[0021] Figure 3 Schematic diagram of the spectrum of single-frequency EEG interference signal. DETAILED DESCRIPTION

[0022] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention 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 invention and are not intended to limit the present invention.

[0023] An interference wave marking algorithm is provided, which is applied to an electroencephalogram and includes a first sub-algorithm and / or a second sub-algorithm. The first sub-algorithm is an instantaneous outlier recognition and processing algorithm for the case where there is an instantaneous large peak interference in the electroencephalogram data. The second sub-algorithm is a single-frequency interference signal recognition and processing algorithm for the case where there is a single-frequency interference signal in the electroencephalogram data.

[0024] Specifically, for EEG data, the first sub-algorithm or the second sub-algorithm can be used for processing; or the first sub-algorithm can be used for processing first to eliminate instantaneous outliers, and then the second sub-algorithm can be used for processing to eliminate single-frequency interference signals; or, the second sub-algorithm can be used for processing first, and then the first sub-algorithm can be used for processing.

[0025] The principle of the instantaneous outlier recognition and processing algorithm is: usually, the brain waves without interference are relatively smooth after filtering. When there is a momentary large peak interference (such as Figure 1 This type of interference can be identified by the amplitude of the deviation from the baseline. The instantaneous outlier identification processing algorithm is as follows:

[0026] S11, dividing the EEG data into a plurality of preset time periods, that is, dividing the total sampling time of the EEG data into a plurality of time periods of equal length, each preset time period including N brain wave signals; in each time period, obtaining the average value of the brain waves in the current time period: , where N is a natural number greater than 1, representing the total amount of data; n represents the nth data.

[0027] S12, in the current time period, the time of the first EEG signal that is greater than or equal to the average value of the EEG and exceeds the preset high threshold is recorded as the interference start time. That is, let S(n) be the value obtained by subtracting the average value of the EEG Vz from the nth EEG signal, and the preset high threshold is a preset instantaneous abnormal peak value, such as 200; if S(n) is greater than or equal to 200, then the time of the EEG signal is the interference start time .

[0028] S13, starting from the interference start time, record the time when the EEG signal is less than or equal to the preset low threshold for M consecutive times and record the interference end time ; Wherein, M is a natural number greater than 1, and the preset lower threshold is a preset abnormal critical value; M is preferably 10, and the lower threshold is preferably 10; when the EEG signal is less than or equal to 10 for 10 consecutive times, it is determined to be the end time of the interference.

[0029] S14, recording the time between the interference start time and the interference end time as the interference time period, that is, to The time period between.

[0030] S15, merging the interference time periods in each preset time period to obtain a first interference time period. That is, repeating steps S12 to S14, performing a union operation on the interference time periods obtained in each preset time period to obtain the first interference time period.

[0031] The principle of the single-frequency interference signal recognition and processing algorithm is as follows: The characteristic of a single-frequency interference signal is that each lead data is affected by the interference signal of the same frequency and amplitude (such as Figure 2 As shown in Figure 2, the Fast Fourier Transform (FFT) is used to analyze the spectrum of the signal every second. At this time, the value of a certain frequency band is abnormally large (such as Figure 3 As shown in Figure 2), if the maximum value of each lead data after FFT conversion is the same as the position where the frequency band appears, then this section of data is interference data, and the time period where the interference data is located is the interference time period. The single-frequency interference signal recognition and processing algorithm includes:

[0032] S21, dividing the EEG data into a plurality of preset time periods, that is, dividing the total sampling time of the EEG data into a plurality of time periods of equal length, each preset time period including a plurality of lead data;

[0033] S22, perform fast Fourier transform on each lead data to obtain the spectrum of the lead data, and each spectrum includes a maximum value; specifically, use fast Fourier transform to analyze the spectrum of the signal with a duration of 1 second in the same time period, and record the maximum value of each lead data. The spectrum at time is: .

[0034] S23, if the maximum value in each spectrum is located at the same position, then the time period where the position is located is the interference time period. The maximum value of appears at position j, and the time period where j is located is the interference time period.

[0035] S24, combining the interference time periods in the plurality of lead data to obtain a second interference time period. That is, repeating steps S22 to S23, performing a union operation on the interference time periods obtained in each preset time period to obtain the second interference time period.

[0036] Furthermore, when the first sub-algorithm and the second sub-algorithm are used successively to process the EEG data, a union operation is performed on the first interference time period and the second interference time period to obtain a total interference time period.

[0037] Furthermore, it also includes an EEG signal acquisition quality evaluation algorithm, which is: the EEG signal acquisition quality is the ratio of the total interference time period to the total sampling time of the EEG data. The EEG signal acquisition quality evaluation algorithm is used to evaluate the quality of the sampled signal and to analyze the weighted judgment of the relevant parameters in the report.

[0038] Specifically, let the total interference duration be , the total sampling time is T, then the interference time proportion p is: When p > 50, it means the signal quality is extremely poor; when 50 > p > 30, it means the signal quality is poor; when 30 > p > 10, it means the signal is average; when P < 10, it means the signal is good.

[0039] A brain fluctuation graph is provided, which adopts the interference wave marking algorithm as described above. Meanwhile, other components and functions of the brain fluctuation graph are not described in detail here.

[0040] The above is an explanation of the interference wave labeling algorithm and brain fluctuation map of the present invention, which is used to help understand the present invention; however, the implementation mode of the present invention is not limited to the above-mentioned embodiments, and any changes, modifications, substitutions, combinations, and simplifications made without departing from the principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. An interference wave marking algorithm, characterized in that: The interference wave labeling algorithm is applied to the brain fluctuation graph, and includes a first sub-algorithm and / or a second sub-algorithm; For the EEG data, use any one of the first sub-algorithm and the second sub-algorithm to process it, or use the first sub-algorithm to process it first and then use the second sub-algorithm to process it, or use the second sub-algorithm to process it first and then use the first sub-algorithm to process it; The first sub-algorithm is an instantaneous outlier identification and processing algorithm, and comprises the steps of: S11, dividing the EEG data into a plurality of preset time periods, each preset time period including N brain wave signals; in each preset time period, obtaining an average value of the brain waves in the current preset time period; wherein N is a natural number greater than 1; S12, within the current preset time period, recording the time at which the first EEG signal greater than or equal to the average value of the EEG wave and exceeding the preset high threshold value occurs as the interference start time; S13, starting from the interference start time, recording the interference end time as the time when the EEG signal is less than or equal to the preset low threshold for M consecutive times; wherein M is a natural number greater than 1; S14, recording the time between the interference start time and the interference end time as the interference time period; S15, combining the interference time periods in each preset time period to obtain a first interference time period; The second sub-algorithm is a single-frequency interference signal identification and processing algorithm, and comprises the steps of: S21, dividing the EEG data into a plurality of preset time periods, each preset time period including a plurality of lead data; S22, performing a fast Fourier transform on each lead data to obtain a frequency spectrum of the lead data, wherein each frequency spectrum includes a maximum value; S23, if the maximum values ​​in each spectrum are located at the same position, the time period at the position is the interference time period; S23, combining interference time periods in the plurality of lead data to obtain a second interference time period.

2. The interference wave marking algorithm as claimed in claim 1, characterized in that: After first using the first sub-algorithm for processing and then using the second sub-algorithm for processing, or after first using the second sub-algorithm for processing and then using the first sub-algorithm for processing, take the union of the first interference time period and the second interference time period to obtain a total interference time period.

3. The interference wave marking algorithm as claimed in claim 2, characterized in that: It also includes an EEG signal acquisition quality evaluation algorithm, which includes: The quality of EEG signal acquisition is the ratio of the total interference time period to the total sampling time of the EEG data.

4. The interference wave marking algorithm according to any one of claims 1 to 3, characterized in that: The high threshold is 200.

5. The interference wave marking algorithm according to any one of claims 1 to 3, characterized in that: The lower threshold is 10.

6. The interference wave marking algorithm according to any one of claims 1 to 3, characterized in that: The M is 10.

7. A brain fluctuation graph, characterized in that: The brain fluctuation graph adopts the interference wave labeling algorithm as described in any one of claims 1 to 6.