Data processing method, device and electronic equipment

Abstract

The present invention relates to the field of brain-computer interface technology, and discloses a data processing method, device and electronic equipment, including: after obtaining an original data set based on an implanted electrode, screening the original data set based on preset conditions to obtain a usable data set, and then classifying it based on a preset format type to obtain a classified data set, which includes: a behavioral data set and a neuron data set, and then comparing the behavioral data set with the neuron data set in time to obtain standard neuron data, filtering and sorting the standard neuron data to obtain neuron sorting data, and finally integrating the data based on specific data standards to obtain a specific data set. Compared with traditional manual management methods, this method can achieve accurate data processing and convenient and fast automated management.

Description

Technical Field

[0001] The present invention belongs to the field of brain-computer interface, and in particular relates to a data processing method, device and electronic equipment. Background Art

[0002] With the progress of the times, brain-computer interface has gradually become a hot field. When electrodes record neural signals in the human or animal brain, a large amount of data will be generated. How to manage this data has become an urgent problem that needs to be solved.

[0003] Traditional data management methods are mainly for experimenters to manually store and manage data, and share data within the laboratory through sharing experimental records or verbal explanations. However, with the development of technology, the number of channels that can be recorded simultaneously in electrophysiological experiments has reached thousands, and the amount of data produced has also increased dramatically. If we continue to use the traditional decentralized storage and management methods, it will take a lot of time and manpower. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a data processing method, device and electronic device, the main purpose of which is to conveniently and quickly automatically manage neural activity data while ensuring the accuracy of data processing results.

[0005] In order to achieve the above object, the present invention mainly provides the following technical solutions: An embodiment of the present invention provides a data processing method, comprising: acquiring an original data set, wherein the original data set is acquired based on implanted electrodes; screening the original data set based on preset conditions to obtain an available data set; classifying the available data set based on a preset format type to obtain a classified data set, wherein the classified data set includes: a behavioral data set and a neuronal data set; performing time comparison between the behavioral data set and the neuronal data set to obtain standard neuronal data; filtering and sorting the standard neuronal data to obtain neuronal sorting data; and performing data integration on the neuronal sorting data based on specific data standards to obtain a specific data set.

[0006] Optionally, after obtaining the classified data set, the method includes: extracting a threshold pass rate and a neural event marker from the neuron data set; converting the format of the neuron data set based on the threshold pass rate and the neural event marker to obtain a converted neuron data set; extracting trial information, and converting the format of the behavior data set based on the trial information to obtain a converted behavior data set; and performing a time comparison between the converted behavior data set and the converted neuron data set.

[0007] Optionally, filtering and sorting the standard neuron data includes: grouping the standard neuron data based on the electrodes to obtain multiple groups of data to be sorted.

[0008] Optionally, after obtaining the plurality of groups of data to be sorted, the method includes: processing the standard neuron data based on a preset sorting algorithm to obtain sorting sub-data; the preset algorithm includes at least: Kilosort 2.5 algorithm and SpykingCircus algorithm.

[0009] Optionally, after obtaining the sorting sub-data, the method includes: screening the sorting sub-data based on preset features to obtain the neuron sorting data; the preset features include at least: waveform, correlation, and amplitude.

[0010] Optionally, the step of integrating the neuron sorting data based on a specific data standard to obtain a specific data set also includes: integrating the neuron sorting data, the standard neuron data, and the behavioral data set based on a specific data standard to obtain the specific data set.

[0011] Optionally, after obtaining the specific data set, it includes: obtaining first electrode channel information and second electrode channel information, wherein the first electrode channel information corresponds to the neuron sorting data, and the second electrode channel information corresponds to the standard neuron data; comparing the first electrode channel information with the second electrode channel information to obtain a first verification result; and / or comparing the behavioral data set with the neuron sorting data to obtain a second verification result.

[0012] Optionally, the filtering of the original data set based on a preset condition to obtain an available data set includes: the preset condition is to meet a preset stored value; and the filtering of the original data based on the preset stored value, if it meets the condition, the data set is available.

[0013] A data processing device, the device comprising: an acquisition unit, a processing unit and an integration unit; the acquisition unit is used to acquire an original data set, wherein the original data set is acquired based on implanted electrodes; the processing unit is used to screen the original data set based on preset conditions to obtain a usable data set; classify the usable data set based on a preset format type to obtain a classified data set, wherein the classified data set includes: a behavioral data set and a neuron data set; perform time comparison between the behavioral data set and the neuron data set to obtain standard neuron data; filter and sort the standard neuron data to obtain neuron sorting data; the integration unit is used to integrate the neuron sorting data based on a specific data standard to obtain a specific data set.

[0014] An electronic device comprises at least one processor, and at least one memory and a bus connected to the processor; wherein the processor and the memory communicate with each other via the bus; and the processor is used to call program instructions in the memory to execute the above-mentioned data processing method.

[0015] The data processing method, device and electronic device provided by the embodiment of the present invention include: after obtaining the original data set based on the implanted electrode, the original data set is screened based on the preset conditions to obtain the available data set, and then classified based on the preset format type to obtain the classified data set, including: behavior data set and neuron data set, and then the behavior data set is compared with the neuron data set in time to obtain standard neuron data, the standard neuron data is filtered and sorted to obtain neuron sorting data, and finally the data is integrated based on the specific data standard to obtain a specific data set. Compared with the traditional manual management method, this method can realize accurate data processing and convenient and fast automatic management. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention provides a flowchart of a data processing method.

[0017] Figure 2 Another flow chart of the data processing method provided by an embodiment of the present invention.

[0018] Figure 3 Another flow chart of the data processing method provided by an embodiment of the present invention.

[0019] Figure 4 A schematic diagram of a data processing device provided by an embodiment of the present invention.

[0020] Figure 5 A schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Please see Figure 1 , Figure 1 A flowchart for implementing the data processing method provided in the embodiment of the present application may include: Step S101: Acquire an original data set, wherein the original data set is acquired based on implanted electrodes.

[0023] Among them, the implanted electrodes can be semi-invasive or invasive, and the original data set can be acquired through multi-channel recording, electrocorticography (ECoG), or stereoelectroencephalography (sEEG). The original data set includes at least: biological neuron data and behavioral data.

[0024] It should be noted that the raw data were obtained through the neural activity recording system and stored in the folder of the corresponding experimental interval (naming rule: date-paradigm-serial number). If there were any problems during the experiment, such as power outages or poor subject status, the experimenter would mark the folder name with a note depending on the possibility of the data being used for subsequent analysis. At the same time, the experimenter also obtained raw behavior-related data through the behavior recording system and motion tracking system.

[0025] Step S102: Filter the original data set based on preset conditions to obtain an available data set.

[0026] The preset condition may be a preset file storage size, a preset file type, a preset file acquisition time, a preset file acquisition duration, a preset file acquisition order, and the like.

[0027] It should be noted that in the process of obtaining the original data set through implanted electrodes, some incomplete data or failed data will appear. These data need to be screened out to avoid affecting the subsequent use or management process. Therefore, they are screened according to preset conditions. If they meet the conditions, they are judged as available data sets. If they do not meet the conditions, they are excluded.

[0028] Step S103: classify the available data sets based on the preset format type to obtain a classified data set, where the classified data set includes: a behavior data set and a neuron data set.

[0029] The available data sets include at least: neuron data and behavior data. The neuron data and behavior data are classified respectively by preset format types to obtain a behavior data set and a neuron data set.

[0030] It should be noted that the raw data sets are classified according to the file format (file name suffix) and stored in specific folders. Since the file formats are unique and fixed, the classification rules are single-mapped and there will be no overlap. For example, files with suffixes of .ns6 or .nev are classified into a folder named "raw_data"; files with suffixes of .bhv2 are classified into a folder named "bhv".

[0031] Step S104: time-matching the behavior data set with the neuron data set to obtain standard neuron data.

[0032] Among them, after comparing the behavior data set with the neuron data set in time, the data that cannot correspond in time are removed. At the same time, the neuron data corresponding to the time of the behavior data is retained, which is called standard neuron data.

[0033] It should be noted that the event times in the behavioral data and neuronal data are extracted from the classification data set respectively, and the difference between the two is subjected to a zero distribution test to determine whether the times of the two data can be aligned. If they can be aligned, they are marked as standard neuronal data.

[0034] Step S105: filter and sort the standard neuron data to obtain neuron sorting data.

[0035] Among them, the standard neuron data is filtered and sorted based on a variety of preset algorithms to obtain neuron sorting data.

[0036] It should be noted that filtering and sorting the standard neuron data to obtain action potential related information can be called data preprocessing, which is divided into four steps: data grouping, rough sorting, result induction, and fine sorting. In other words, data grouping is first performed, and then rough sorting is performed using at least two algorithms, and then the results of the rough sorting are combined to modify the data labels in the main preliminary sorting data set to obtain sorted sub-data, and then the sorted sub-data are finely sorted.

[0037] Step S106: integrating the neuron sorting data based on a specific data standard to obtain a specific data set.

[0038] The specific data standard is a specific data conversion template. The neuron sorting data is placed in the specific data conversion template to form data in the .nwb file format, which is then integrated and stored.

[0039] It should be noted that only the neuron sorting data can be put into a specific data conversion template for integration; or the neuron sorting data, standard neuron data and behavioral data can be integrated (i.e., the data before and after filtering and sorting are retained at the same time) and converted into a .nwb file to obtain a specific data set. In other words, the specific data set can include only the neuron data after integration, or it can include the neuron data after sorting, the neuron data before sorting and the behavioral data.

[0040] Please see Figure 2 , the method of this embodiment may include: after obtaining the classification data set, including: Step S201: extracting threshold passing rate and neural event markers from the neuron data set; Step S202: converting the format of the neuron dataset based on the threshold passing rate and the neural event marker to obtain a converted neuron dataset; Step S203: extracting trial information, and converting the format of the behavior dataset based on the trial information to obtain a converted behavior dataset.

[0041] Step S204: time-matching the converted behavior dataset with the converted neuron dataset.

[0042] Among them, the classification data includes both behavioral data sets and neuronal data sets; the extraction of threshold pass rate and neural event markers is only performed on neuronal data sets, while the extraction of trial information is only performed on behavioral data sets. At the same time, trial information refers to: visual stimulation parameters and behavioral parameters that appear in each trial. In the entire trial, it can be roughly divided into several key events: start-visual stimulation-movement-end. Neural event markers are to record the event sequence number and the corresponding timestamp.

[0043] It should be noted that the execution order of steps S202 and S203 is not restricted and can be switched with each other, but step S202 needs to be completed based on S201. The key point is how to obtain the converted behavior dataset and the converted neuron dataset, and perform time comparison.

[0044] The method of this embodiment may include: filtering and sorting the standard neuron data, including: grouping the standard neuron data based on electrodes to obtain multiple groups of data to be sorted.

[0045] It should be noted that the grouping of standard neuron data based on electrodes, that is, the data grouping in the four steps of data preprocessing mentioned above, is to group the standard neuron data according to the electrode site distribution information, so as to obtain data units corresponding to each sector (shank) of the electrode, and these data units constitute a data set.

[0046] The method of this embodiment may include: after obtaining multiple groups of data to be sorted, including: processing the standard neuron data based on a preset sorting algorithm to obtain sorting sub-data; the preset algorithm includes at least: Kilosort 2.5 algorithm and SpykingCircus algorithm.

[0047] It should be noted that the standard neuron data is processed based on the preset sorting algorithm, that is, the rough sorting in the four steps of data preprocessing mentioned above, and each data unit in the sorting data set is automatically sorted using at least two algorithms (which can be Kilosort 2.5 and SpykingCircus or others) to obtain the initial sorting data set and the reference data set. The results of the reference data set are combined to automatically modify the data labels in the initial sorting data set to obtain the sorted sub-data.

[0048] The method of this embodiment may include: after obtaining the sorting sub-data, including: screening the sorting sub-data based on preset features to obtain neuron sorting data; the preset features include at least: waveform, correlation, amplitude.

[0049] It should be noted that the neuron sorting data is screened based on preset features, that is, the fine sorting in the four steps of data preprocessing mentioned above is based on factors such as waveform, correlation, voltage amplitude, etc., to check and confirm the sorting sub-data.

[0050] The method of this embodiment may include: integrating neuron sorting data based on specific data standards to obtain a specific data set, and also includes: integrating sorting data, standard neuron data, and behavioral data sets based on specific data standards to obtain a specific data set.

[0051] It should be noted that a variety of different types of data can be integrated to form a specific data set, that is, including: sorting data, standard neuron data, and behavioral data set.

[0052] Please see Figure 3 , the method of this embodiment may include: after obtaining a specific data set, including: Step S301: acquiring first electrode channel information and second electrode channel information, wherein the first electrode channel information corresponds to neuron sorting data, and the second electrode channel information corresponds to standard neuron data; Step S302: comparing the first electrode channel information with the second electrode channel information to obtain a first verification result; and / or comparing the behavior data set with the neuron sorting data to obtain a second verification result.

[0053] After obtaining a specific data set, an evaluation may be performed based on the first verification result and the second verification result to ensure the reliability of the data.

[0054] It should be noted that the verification of integrated data can be divided into three items, namely: (1) Temporal consistency: Temporal alignment test of behavioral datasets and neuron sorting data.

[0055] (2) Channel consistency: Perform consistency check on neuron sorting data and standard neuron data, that is, whether the electrode channel information of the two can be verified with each other.

[0056] (3) Activity specificity: Based on the behavioral data set, the neuron sorting data is analyzed and tested to determine whether the neural activity patterns in the neuron sorting data have a correlation with the behavioral events that is consistent with biological findings.

[0057] If the check passes, the final verification result is obtained.

[0058] The method of this embodiment may include: a preset condition is to meet a preset storage value; based on the preset storage value, the original data is screened, and if it meets the condition, it is an available data set.

[0059] It should be noted that the preset storage value can be a value, a set of values, or an interval. For example: determine whether the size of each data in the original data set is greater than 50k, if it is greater, it meets the requirements, if it is less, it is eliminated; another example: determine whether the size of each data in the original data set is between 30k-40k, if it is within this range, it meets the requirements, if it is not within this range, it is eliminated.

[0060] Please see Figure 4 , the device of this embodiment may include the following structure: An acquisition unit 401 is used to acquire an original data set, wherein the original data set is acquired based on an implanted electrode; The processing unit 402 is used to filter the original data set based on preset conditions to obtain an available data set; classify the available data set based on a preset format type to obtain a classified data set, wherein the classified data set includes a behavior data set and a neuron data set; perform time comparison between the behavior data set and the neuron data set to obtain standard neuron data; filter and sort the standard neuron data to obtain neuron sorting data; The integration unit 403 is used to integrate the neuron sorting data based on a specific data standard to obtain a specific data set.

[0061] Please see Figure 5 This embodiment of the present application also discloses an electronic device, which includes at least one processor 501, and at least one memory 502 and a bus 503 connected to the processor 501; wherein the processor 501 and the memory 502 communicate with each other through the bus 503; the processor 501 is used to call the program instructions in the memory 502 to execute the above-mentioned data processing method based on the invasive brain-computer interface.

[0062] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A data processing method, characterized in that: include: Acquire an original data set, wherein the original data set is acquired based on an implanted electrode; Screening the original data set based on preset conditions to obtain an available data set; Classifying the available data sets based on a preset format type to obtain a classified data set, wherein the classified data set includes: a behavior data set and a neuron data set; Performing time comparison between the behavior dataset and the neuron dataset to obtain standard neuron data; Filtering and sorting the standard neuron data to obtain neuron sorting data; The neuron sorting data is integrated based on a specific data standard to obtain a specific data set.

2. The processing method according to claim 1, characterized in that: After obtaining the classification data set, the method includes: extracting threshold passing rate and neural event markers from the neuron data set; Performing format conversion on the neuron dataset based on the threshold passing rate and the neural event marker to obtain a converted neuron dataset; Extracting trial information, and converting the format of the behavior data set based on the trial information to obtain a converted behavior data set; The transformed behavioral dataset is temporally aligned with the transformed neuronal dataset.

3. The processing method according to claim 1, characterized in that: Filtering and sorting the standard neuron data includes: The standard neuron data are grouped based on the electrodes to obtain multiple groups of data to be sorted.

4. The processing method according to claim 3, characterized in that: After obtaining multiple groups of data to be sorted, the method includes: Processing the standard neuron data based on a preset sorting algorithm to obtain sorting sub-data; The preset algorithms include at least: Kilosort 2.5 algorithm and SpykingCircus algorithm.

5. The processing method according to claim 4, characterized in that: After obtaining the sorting sub-data, the following steps are included: Screening the sorting sub-data based on preset features to obtain the neuron sorting data; The preset features include at least: waveform, correlation, and amplitude.

6. The processing method according to claim 5, characterized in that: The step of integrating the neuron sorting data based on a specific data standard to obtain a specific data set further includes: The neuron sorting data, the standard neuron data, and the behavior data set are integrated based on a specific data standard to obtain the specific data set.

7. The processing method according to claim 6, characterized in that: After obtaining the specific data set, the method includes: Acquire first electrode channel information and second electrode channel information, wherein the first electrode channel information corresponds to the neuron sorting data, and the second electrode channel information corresponds to the standard neuron data; The first electrode channel information is compared with the second electrode channel information to obtain a first verification result; And / or, the behavior data set is compared with the neuron sorting data to obtain a second verification result.

8. The processing method according to claim 1, characterized in that: The screening of the original data set based on the preset conditions to obtain the available data set includes: The preset condition is to meet the preset storage value; The original data is screened based on the preset storage value, and if it meets the requirements, it is an available data set.

9. A data processing device, characterized in that: The device comprises: an acquisition unit, a processing unit and an integration unit; The acquisition unit is used to acquire an original data set, wherein the original data set is acquired based on implanted electrodes; The processing unit is used to screen the original data set based on preset conditions to obtain an available data set; classify the available data set based on a preset format type to obtain a classified data set, wherein the classified data set includes a behavior data set and a neuron data set; perform time comparison between the behavior data set and the neuron data set to obtain standard neuron data; filter and sort the standard neuron data to obtain neuron sorting data; The integration unit is used to integrate the neuron sorting data based on a specific data standard to obtain a specific data set.

10. An electronic device, characterized in that: The electronic device includes at least one processor, and at least one memory and a bus connected to the processor; wherein the processor and the memory communicate with each other through the bus; the processor is used to call program instructions in the memory to execute the data processing method described in any one of claims 1-8.

Images