Electroencephalogram data timestamp quality evaluation and restoration method and quality evaluation and restoration system

Abstract

The invention discloses an electroencephalogram data timestamp quality evaluation and restoration method and a quality evaluation and restoration system, and belongs to the field of industrial internet. According to the method and the device, the problem of poor accuracy and reliability of the electroencephalogram data caused by asynchronous timestamps of the electroencephalogram data in the prior art is solved. According to the method, the evaluation standard of the timestamp quality is provided by calculating the consistency value, redundant data can be identified and removed, and the validity and the accuracy of the data are improved. The accuracy of arrival of the data points according to the expected time is measured by calculating the timeliness value, the influence of data delay on analysis is reduced, and the real-time performance of data is improved. The timestamp sequence is reconstructed based on the dynamic programming algorithm, and the continuity, reliability and overall quality of the data can be improved while the repair efficiency is improved with the purpose of minimizing the repair cost, so that the method is more suitable for timestamp repair of large-scale data. The method can be applied to quality evaluation and restoration of the electroencephalogram data timestamps.

Description

Technical Field

[0001] The present invention belongs to the field of industrial Internet, and specifically relates to an electroencephalogram data timestamp quality assessment and repair method and a quality assessment and repair system. Background Art

[0002] With the rapid development of brain science and neural engineering, electroencephalography (EEG) as a non-invasive means of monitoring brain activity has been widely used in clinical diagnosis, brain-computer interface and cognitive science research. The quality of EEG data directly affects the effect of research and application.

[0003] However, the existing EEG data collection and transmission process has the following technical defects:

[0004] (1) Device delay: The processing speeds of different devices vary, resulting in timestamp asynchrony.

[0005] (2) Sampling rate asynchrony: The sampling rates of different devices are inconsistent or unstable, resulting in inaccurate data.

[0006] (3) Data transmission delay: Especially in the case of wireless transmission, the delay in signal transmission affects the accuracy of the timestamp.

[0007] In summary, the above defects can lead to problems such as asynchrony of EEG data timestamps, affecting the accuracy and reliability of EEG data. Summary of the invention

[0008] The purpose of the present invention is to solve the problem in the prior art that the accuracy and reliability of EEG data are poor due to the asynchrony of the timestamp of the EEG data, and to propose an EEG data timestamp quality assessment and repair method and a quality assessment and repair system.

[0009] The technical solution adopted by the present invention to solve the above technical problems is:

[0010] According to one aspect of the present invention, a method for evaluating and repairing the timestamp quality of EEG data is provided, and the method specifically comprises the following steps:

[0011] Step 1: Collect EEG data using electrodes placed on the surface of the scalp;

[0012] Step 2: Calculate the consistency and timeliness of the timestamp sequence corresponding to the collected EEG data;

[0013] Step 3: repair the timestamp sequence corresponding to the EEG data according to the calculation result of step 2;

[0014] Step 4: Store the EEG data with the repaired timestamp in a set data format.

[0015] According to another aspect of the present invention, a system for evaluating and repairing the timestamp quality of EEG data is provided, the system comprising an EEG data acquisition module, a timestamp sequence consistency calculation module, a timestamp sequence timeliness calculation module and a timestamp sequence repair module;

[0016] An EEG data acquisition module is used to collect EEG data using electrodes arranged on the surface of the scalp;

[0017] A timestamp sequence consistency calculation module is used to calculate the consistency of the timestamp sequence corresponding to the collected EEG data;

[0018] The consistency calculation method is:

[0019]

[0020] Among them, consistency represents the consistency of the timestamp sequence, redundancyCnt represents the redundant timestamp count, and Cnt represents the total timestamp count;

[0021] A timestamp sequence timeliness calculation module is used to calculate the timeliness of the timestamp sequence corresponding to the collected EEG data;

[0022] The timeliness calculation method is:

[0023]

[0024] Among them, timeliness represents the timeliness of the timestamp sequence, and lateCnt represents the delayed timestamp count;

[0025] The timestamp sequence repair module is used to repair the timestamp sequence according to the consistency and timeliness calculation results:

[0026] Compare consistency with threshold θ consistency The size of timeliness is compared with the threshold θ timeliness Size:

[0027] If consistency is less than the threshold θ consistency Or timeliness is less than the threshold θ timeliness , then the quality of the timestamp sequence is unqualified, and the timestamp sequence corresponding to the EEG data needs to be repaired;

[0028] Otherwise, the quality of the timestamp sequence is qualified, and there is no need to repair the timestamp sequence corresponding to the EEG data.

[0029] The beneficial effects of the present invention are:

[0030] The present invention provides an evaluation standard for timestamp quality by calculating consistency values, helps identify and remove redundant data, and improves the validity and accuracy of data. By calculating the timeliness value to measure the accuracy of data points arriving at the expected time, the impact of data delay on analysis is reduced, and the real-time nature of data is improved. Reconstructing the timestamp sequence based on the dynamic programming algorithm with the goal of minimizing the repair cost can improve the continuity, reliability and overall quality of data while improving the repair efficiency, and is more suitable for timestamp repair of large-scale data.

[0031] Moreover, the EEG data of the present invention is stored based on a time series database, which facilitates data management, query and analysis and improves the efficiency and reliability of data storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of a method for evaluating and repairing the timestamp quality of EEG data of the present invention. DETAILED DESCRIPTION

[0033] Specific implementation method 1: Combination Figure 1 The present embodiment describes a method for evaluating and repairing the timestamp quality of electroencephalogram data, and the method specifically comprises the following steps:

[0034] Step 1: Collect EEG data using electrodes placed on the surface of the scalp;

[0035] Step 2: Calculate the consistency and timeliness of the timestamp sequence corresponding to the collected EEG data (the EEG data of each channel corresponds to the same timestamp sequence);

[0036] Step 3: repair the timestamp sequence corresponding to the EEG data according to the calculation result of step 2;

[0037] Step 4: Store the EEG data with the repaired timestamp in a set data format.

[0038] The present invention stores data based on a time series database, and the stored data format is: timestamp, experiment ID, date, channel 1 signal value, channel 2 signal value, ..., channel n signal value, initial timestamp.

[0039] Specific implementation method 2: This implementation method is different from the specific implementation method 1 in that the step 1 is specifically as follows:

[0040] Step 1: Determine the placement of electrodes on the scalp surface;

[0041] Step 12: Arrange the electrodes made of metal at the positions determined in step 11, and place conductive paste between the electrodes and the scalp to enhance signal conduction;

[0042] Step 1: Each electrode is connected to the EEG amplifier through a wire, and the amplified EEG data of each channel is recorded.

[0043] The other steps and parameters are the same as those in the first embodiment.

[0044] Due to the following reasons, the timestamp interval may be out of sync during the data collection process.

[0045] Equipment delays: Different processing speeds between different devices (such as EEG amplifiers, computers, etc.) may cause timestamps to become out of sync.

[0046] Different sampling rates: If the sampling rates of different devices are not synchronized, or the sampling rates of some devices are unstable, it will cause inconsistent timestamps.

[0047] Data transmission delay: The signals collected by the electrodes are transmitted to the amplifier and computer via wires. There may be delays in the transmission process, especially when wireless transmission is used.

[0048] Therefore, timestamp quality assessment and repair are needed.

[0049] Specific implementation method three: This implementation method is different from specific implementation methods one or two in that the specific process of step two is as follows:

[0050] Calculate the consistency of the timestamp sequence corresponding to the collected EEG data:

[0051]

[0052] Among them, consistency represents the consistency of the timestamp sequence, redundancyCnt represents the redundant timestamp count, and Cnt represents the total timestamp count;

[0053] Calculate the timeliness of the timestamp sequence corresponding to the collected EEG data:

[0054]

[0055] Among them, timeliness represents the timeliness of the timestamp sequence, and lateCnt represents the delayed timestamp count.

[0056] The other steps and parameters are the same as those in the first or second embodiment.

[0057] Consistency indicates the proportion of redundant data points in a data set, and timeliness is used to measure whether data points arrive at the expected time interval. The present invention evaluates the quality of timestamps by evaluating the consistency and timeliness of timestamp sequences, thereby improving the accuracy of subsequent data analysis and decision-making.

[0058] Specific implementation method 4: This implementation method is different from any one of the specific implementation methods 1 to 3 in that the specific process of step 3 is as follows:

[0059] Compare consistency with threshold θ consistency (customizable) size, compare timeliness with threshold θ timeliness (Customizable) size:

[0060] If consistency is less than the threshold θ consistency Or timeliness is less than the threshold θ timeliness , then the quality of the timestamp sequence is unqualified, and the timestamp sequence corresponding to the EEG data needs to be repaired;

[0061] Otherwise, the quality of the timestamp sequence is qualified, and there is no need to repair the timestamp sequence corresponding to the EEG data.

[0062] The other steps and parameters are the same as those in Specific Embodiments 1 to 3.

[0063] Specific implementation method 5: This implementation method is different from any one of the specific implementation methods 1 to 4 in that the timestamp sequence corresponding to the EEG data is repaired, and the specific process is as follows:

[0064] Step 1, calculate the standard timestamp interval interval and the starting timestamp startTime;

[0065] Step 2: Repair the timestamp sequence based on the dynamic programming algorithm and the calculated standard timestamp interval and start timestamp.

[0066] The other steps and parameters are the same as those in Specific Embodiments 1 to 4.

[0067] The objective function of the dynamic programming algorithm is to minimize the repair cost. The operations that generate the repair cost include move, insert, and delete. Construct a dynamic programming table DP and calculate the table DP through the dynamic programming state transition equation. The dynamic programming state transition equation is:

[0068] DP(i,j)=min{DP(i-1,j-1)+Δ m ,DP(i,j-1)+Δ a , DP(i-1, j)+Δ d}

[0069] Where DP(i-1, j-1) is the minimum repair cost of the first j-1 timestamps in the repaired sequence relative to the first i-1 timestamps in the unrepaired sequence, that is, the element in the i-1th row and j-1th column in table DP; DP(i, j-1) is the minimum repair cost of the first j-1 timestamps in the repaired sequence relative to the first i timestamps in the unrepaired sequence; DP(i-1, j) is the minimum repair cost of the first j timestamps in the repaired sequence relative to the first i-1 timestamps in the unrepaired sequence; Δ m , Δ a and Δ d They are move cost, insertion cost and deletion cost respectively.

[0070] After the calculation is completed, the best timestamp repair sequence is determined by backtracking the filled matrix.

[0071] Specific implementation method 6: This implementation method is different from any one of specific implementation methods 1 to 5 in that the calculation method used to calculate the standard timestamp interval interval is method 1, method 2 or method 3:

[0072] Method 1: Calculate the time interval between every two adjacent timestamps, and then take the median of the time interval as the standard timestamp interval;

[0073] Method 2: Calculate the time interval between every two adjacent timestamps, and then take the mode of the time interval as the standard timestamp interval;

[0074] Method 3: Calculate the time interval between every two adjacent timestamps, then use the K-means clustering algorithm to cluster all time intervals, and take the mean of all time intervals in the largest cluster as the standard timestamp interval.

[0075] The other steps and parameters are the same as those in Specific Implementations 1 to 5.

[0076] Method 1 and method 2 have faster calculation speeds. When the amount of data is large, the median or mode calculation method should be used.

[0077] Specific implementation method 7: This implementation method is different from any one of specific implementation methods 1 to 6 in that the calculation method used to calculate the start timestamp startTime is method 1 or method 2:

[0078] Method 1: Linear regression

[0079] Step S1, record the total number of all timestamps as Q, and calculate the sum Q′ of all Q timestamps;

[0080] Step S2, calculate the product of Q-1 and interval, and then subtract Q′ from the product to obtain the subtraction result;

[0081] Step S3, divide the difference result by the number of timestamps Q to obtain the starting timestamp startTime;

[0082] Method 2: Mode of Modulus

[0083] Step P1, respectively calculate the remainder of each timestamp divided by the standard timestamp interval;

[0084] Step P2, count the number of times each remainder appears, and select the majority of the remainders;

[0085] Step P3, find the first timestamp that matches the majority as the temporary starting timestamp startTime;

[0086] Step P4: Calculate the minimum positive integer a that satisfies "temporary start timestamp - a·interval is less than the first timestamp", and then use "temporary start timestamp - a·interval" as the final start timestamp startTime.

[0087] The other steps and parameters are the same as those in Specific Embodiments 1 to 6.

[0088] Specific implementation eight: A system for evaluating and repairing EEG data timestamp quality described in this implementation includes an EEG data acquisition module, a timestamp sequence consistency calculation module, a timestamp sequence timeliness calculation module, and a timestamp sequence repair module;

[0089] An EEG data acquisition module is used to collect EEG data using electrodes arranged on the surface of the scalp;

[0090] A timestamp sequence consistency calculation module is used to calculate the consistency of the timestamp sequence corresponding to the collected EEG data;

[0091] The consistency calculation method is:

[0092]

[0093] Among them, consistency represents the consistency of the timestamp sequence, redundancyCnt represents the redundant timestamp count, and Cnt represents the total timestamp count;

[0094] A timestamp sequence timeliness calculation module is used to calculate the timeliness of the timestamp sequence corresponding to the collected EEG data;

[0095] The timeliness calculation method is:

[0096]

[0097] Among them, timeliness represents the timeliness of the timestamp sequence, and lateCnt represents the delayed timestamp count;

[0098] The timestamp sequence repair module is used to repair the timestamp sequence according to the consistency and timeliness calculation results:

[0099] Compare consistency with threshold θ consistency (customizable) size, compare timeliness with threshold θ timeliness (Customizable) size:

[0100] If consistency is less than the threshold θ consistency Or timeliness is less than the threshold θ timeliness , then the quality of the timestamp sequence is unqualified, and the timestamp sequence corresponding to the EEG data needs to be repaired;

[0101] Otherwise, the quality of the timestamp sequence is qualified, and there is no need to repair the timestamp sequence corresponding to the EEG data.

[0102] Specific implementation method 9: This implementation method is different from specific implementation method 8 in that the specific process of repairing the timestamp sequence corresponding to the EEG data is as follows:

[0103] Step 1: Use method 1, method 2, or method 3 to calculate the standard timestamp interval:

[0104] Method 1: Calculate the time interval between every two adjacent timestamps, and then take the median of the time interval as the standard timestamp interval;

[0105] Method 2: Calculate the time interval between every two adjacent timestamps, and then take the mode of the time interval as the standard timestamp interval;

[0106] Method 3: Calculate the time interval between every two adjacent timestamps, and then use the K-means clustering algorithm to cluster all the time intervals, and take the mean of all the time intervals in the largest cluster as the standard timestamp interval;

[0107] Step 2: Use method 1 or method 2 to calculate the starting timestamp startTime:

[0108] Method 1: Linear regression

[0109] Step S1, record the total number of all timestamps as Q, and calculate the sum Q′ of all Q timestamps;

[0110] Step S2, calculate the product of Q-1 and interval, and then subtract Q′ from the product to obtain the subtraction result;

[0111] Step S3, divide the difference result by the number of timestamps Q to obtain the starting timestamp startTime;

[0112] Method 2: Mode of Modulus

[0113] Step P1, respectively calculate the remainder of each timestamp divided by the standard timestamp interval;

[0114] Step P2, count the number of times each remainder appears, and select the majority of the remainders;

[0115] Step P3, find the first timestamp that matches the majority as the temporary starting timestamp startTime;

[0116] Step P4, calculate the minimum positive integer a that satisfies "temporary start timestamp - a·interval is less than the first timestamp", and use "temporary start timestamp - a·interval" as the final start timestamp startTime;

[0117] Step 3: Repair the timestamp sequence based on the dynamic programming algorithm and the calculated standard timestamp interval and start timestamp.

[0118] The other steps and parameters are the same as those in the eighth embodiment.

[0119] Specific implementation method ten: This implementation method is different from specific implementation methods eight or nine in that the system further includes a data storage module, and the data storage module stores the EEG data with the repaired timestamp according to a set data format.

[0120] The other steps and parameters are the same as those in the eighth or ninth embodiment.

[0121] The above calculation examples of the present invention are only used to explain the calculation model and calculation process of the present invention in detail, and are not intended to limit the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for evaluating and repairing the timestamp quality of EEG data, characterized in that: The method specifically comprises the following steps: Step 1: Collect EEG data using electrodes placed on the surface of the scalp; Step 2: Calculate the consistency and timeliness of the timestamp sequence corresponding to the collected EEG data; Step 3: repair the timestamp sequence corresponding to the EEG data according to the calculation result of step 2; Step 4: Store the EEG data with the repaired timestamp in a set data format.

2. A method for evaluating and repairing EEG data timestamp quality according to claim 1, characterized in that: The step 1 is specifically as follows: Step 1: Determine the placement of electrodes on the scalp surface; Step 12: Arrange the electrodes made of metal at the positions determined in step 11, and place conductive paste between the electrodes and the scalp; Step 1: Each electrode is connected to the EEG amplifier through a wire, and the amplified EEG data of each channel is recorded.

3. The method for evaluating and repairing EEG data timestamp quality according to claim 1, characterized in that: The specific process of step 2 is as follows: Calculate the consistency of the timestamp sequence corresponding to the collected EEG data: Among them, consistency represents the consistency of the timestamp sequence, redundancyCnt represents the redundant timestamp count, and Cnt represents the total timestamp count; Calculate the timeliness of the timestamp sequence corresponding to the collected EEG data: Among them, timeliness represents the timeliness of the timestamp sequence, and lateCnt represents the delayed timestamp count.

4. A method for evaluating and repairing EEG data timestamp quality according to claim 3, characterized in that: The specific process of step three is: Compare consistency with threshold θ consistency The size of timeliness is compared with the threshold θ timeliness Size: If consistency is less than the threshold θ consistency Or timeliness is less than the threshold θ timeliness , then the quality of the timestamp sequence is unqualified, and the timestamp sequence corresponding to the EEG data needs to be repaired; Otherwise, the quality of the timestamp sequence is qualified, and there is no need to repair the timestamp sequence corresponding to the EEG data.

5. A method for evaluating and repairing EEG data timestamp quality according to claim 4, characterized in that: The specific process of repairing the timestamp sequence corresponding to the EEG data is as follows: Step 1, calculate the standard timestamp interval interval and the starting timestamp startTime; Step 2: Repair the timestamp sequence based on the dynamic programming algorithm and the calculated standard timestamp interval and start timestamp.

6. A method for evaluating and repairing EEG data timestamp quality according to claim 5, characterized in that: The calculation method used to calculate the standard timestamp interval interval is method 1, method 2 or method 3: Method 1: Calculate the time interval between every two adjacent timestamps, and then take the median of the time interval as the standard timestamp interval; Method 2: Calculate the time interval between every two adjacent timestamps, and then take the mode of the time interval as the standard timestamp interval; Method 3: Calculate the time interval between every two adjacent timestamps, then use the K-means clustering algorithm to cluster all time intervals, and take the mean of all time intervals in the largest cluster as the standard timestamp interval.

7. A method for evaluating and repairing EEG data timestamp quality according to claim 6, characterized in that: The calculation method used for calculating the start timestamp startTime is method 1 or method 2: Method 1: Step S1, record the total number of all timestamps as Q, and calculate the sum Q′ of all Q timestamps; Step S2, calculate the product of Q-1 and interval, and then subtract Q′ from the product to obtain the subtraction result; Step S3, divide the difference result by the number of timestamps Q to obtain the starting timestamp startTime; Method 2: Step P1, respectively calculate the remainder of each timestamp divided by the standard timestamp interval; Step P2, count the number of times each remainder appears, and select the majority of the remainders; Step P3, find the first timestamp that matches the majority as the temporary starting timestamp startTime; Step P4, calculate the minimum positive integer a that satisfies "temporary start timestamp - a·interval is less than the first timestamp", and then use "temporary start timestamp - a·interval" as the final start timestamp startTime.

8. A system for evaluating and repairing the timestamp quality of electroencephalogram data, characterized in that: The system includes an EEG data acquisition module, a timestamp sequence consistency calculation module, a timestamp sequence timeliness calculation module and a timestamp sequence repair module; An EEG data acquisition module is used to collect EEG data using electrodes arranged on the surface of the scalp; A timestamp sequence consistency calculation module is used to calculate the consistency of the timestamp sequence corresponding to the collected EEG data; The consistency calculation method is: Among them, consistency represents the consistency of the timestamp sequence, redundancyCnt represents the redundant timestamp count, and Cnt represents the total timestamp count; A timestamp sequence timeliness calculation module is used to calculate the timeliness of the timestamp sequence corresponding to the collected EEG data; The timeliness calculation method is: Among them, timeliness represents the timeliness of the timestamp sequence, and lateCnt represents the delayed timestamp count; The timestamp sequence repair module is used to repair the timestamp sequence according to the consistency and timeliness calculation results: Compare consistency with threshold θ consistency The size of timeliness is compared with the threshold θ timeliness Size: If consistency is less than the threshold θ consistency Or timeliness is less than the threshold θ timeliness , then the quality of the timestamp sequence is unqualified, and the timestamp sequence corresponding to the EEG data needs to be repaired; Otherwise, the quality of the timestamp sequence is qualified, and there is no need to repair the timestamp sequence corresponding to the EEG data.

9. The EEG data timestamp quality assessment and repair system according to claim 8, characterized in that: The specific process of repairing the timestamp sequence corresponding to the EEG data is as follows: Step 1: Use method 1, method 2, or method 3 to calculate the standard timestamp interval: Method 1: Calculate the time interval between every two adjacent timestamps, and then take the median of the time interval as the standard timestamp interval; Method 2: Calculate the time interval between every two adjacent timestamps, and then take the mode of the time interval as the standard timestamp interval; Method 3: Calculate the time interval between every two adjacent timestamps, and then use the K-means clustering algorithm to cluster all the time intervals, and take the mean of all the time intervals in the largest cluster as the standard timestamp interval; Step 2: Use method 1 or method 2 to calculate the starting timestamp startTime: Method 1: Step S1, record the total number of all timestamps as Q, and calculate the sum Q′ of all Q timestamps; Step S2, calculate the product of Q-1 and interval, and then subtract Q′ from the product to obtain the subtraction result; Step S3, divide the difference result by the number of timestamps Q to obtain the starting timestamp startTime; Method 2: Step P1, respectively calculate the remainder of each timestamp divided by the standard timestamp interval; Step P2, count the number of times each remainder appears, and select the majority of the remainders; Step P3, find the first timestamp that matches the majority as the temporary starting timestamp startTime; Step P4, calculate the minimum positive integer a that satisfies "temporary start timestamp - a interval is less than the first timestamp", and use "temporary start timestamp - a interval" as the final start timestamp startTime; Step 3: Repair the timestamp sequence based on the dynamic programming algorithm and the calculated standard timestamp interval and start timestamp.

10. The EEG data timestamp quality assessment and repair system according to claim 9, characterized in that: The system also includes a data storage module, which stores the EEG data after the timestamp is repaired according to a set data format.

Images