Information processing method, information processing device and computer program

Abstract

To provide an information processing method that provides a storage area for storing brain data for an account of a user, and can improve convenience of storage and data analysis of brain data, taking into consideration a user who is not a specialist in data processing.SOLUTION: An information processing method that provides a storage area for storing brain data for an account of a user comprises: storing uploaded brain data in a storage in association with the account of the user; acquiring metadata associated with the brain data; and storing the acquired metadata in association with the brain data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing method, an information processing device, and a computer program. [Background technology]

[0002] There is a computing device that includes a data store that stores time-series waveform data and analyzes the time-series waveform data (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Special Publication No. 2015-536170 Summary of the Invention [Problem to be solved by the invention]

[0004] An object of the present disclosure is to provide an information processing method, an information processing device, and a computer program that provide a storage area for storing brain data for a user's account, and that take into consideration users who are not experts in data processing and can improve the convenience of storing and analyzing brain data. [Means for solving the problem]

[0005] An information processing method according to one aspect of the present disclosure is an information processing method that provides a storage area for storing brain data for a user's account, associates uploaded brain data with the user's account and stores it in storage, acquires metadata related to the brain data, and stores the acquired metadata in association with the brain data.

[0006] An information processing device according to one aspect of the present disclosure is an information processing device that provides a storage area for storing brain data for a user's account, and includes: a storage that stores uploaded brain data in association with the user's account; a communication unit that receives metadata related to the brain data; and a processing unit that stores the metadata received by the communication unit in association with the brain data.

[0007] A computer program according to one aspect of the present disclosure is an information processing method for providing a storage area for storing brain data for a user's account, and causes a computer to execute a process of storing uploaded brain data in storage in association with the user's account, acquiring metadata related to the brain data, and storing the acquired metadata in association with the brain data. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to provide a storage area for storing brain data for a user's account, and to improve the convenience of storing and analyzing brain data by taking into consideration users who are not experts in data processing. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a conceptual diagram showing the configuration of an information processing system according to a first embodiment. [Figure 2] 2 is a block diagram showing an example of the configuration of a cloud server according to the first embodiment. FIG. [Figure 3] 10 is a flowchart showing a processing procedure for storing brain data, labeling, etc. [Figure 4] FIG. 1 is a conceptual diagram showing a method for storing brain data. [Figure 5] FIG. 1 is a conceptual diagram showing brain data etc. stored in storage. [Figure 6] 10 is a flowchart showing a processing procedure for data analysis using brain data stored in storage. [Figure 7]FIG. 1 is a conceptual diagram showing a data analysis processing method using brain data stored in storage. [Figure 8] 10 is a flowchart showing a processing procedure for uploading brain data, analyzing the data, and transferring and saving the data. [Figure 9] 10 is a flowchart showing a processing procedure for uploading brain data, analyzing the data, and transferring and saving the data. [Figure 10] FIG. 1 is a conceptual diagram showing a data analysis processing method performed by uploading brain data. [Figure 11] FIG. 1 is a conceptual diagram showing a method for transferring and saving brain data to be analyzed in storage. [Figure 12] 10 is a flowchart showing the processing procedure for storing brain data, adding metadata, etc. according to the second embodiment. [Figure 13] FIG. 10 is a schematic diagram showing an example of a metadata input screen during uploading of brain data according to the second embodiment. [Figure 14] FIG. 10 is a schematic diagram showing an example of a metadata input screen during uploading of brain data according to the second embodiment. [Figure 15] FIG. 10 is a schematic diagram showing an example of a brain data analysis screen according to the second embodiment. [Figure 16] FIG. 10 is a schematic diagram showing a state in which an analysis data tab in the workspace is selected. [Figure 17] FIG. 10 is a schematic diagram showing a state in which a data content tab of a workspace is selected. [Figure 18] 10 is a flowchart showing a processing procedure for data analysis according to the third embodiment using brain data stored in a storage. [Figure 19] FIG. 10 is a conceptual diagram showing a data analysis processing method according to a third embodiment using brain data stored in a storage. [Figure 20] 10 is a flowchart showing a processing procedure for uploading brain data, analyzing the data, and transferring and saving the data according to the third embodiment. [Figure 21] FIG. 10 is a conceptual diagram showing a data analysis processing method according to a third embodiment, which is performed by uploading brain data. [Figure 22] FIG. 11 is a conceptual diagram showing a method for transferring and saving brain data to be analyzed in a storage device according to the third embodiment. [Figure 23] FIG. 11 is a schematic diagram showing a folder list screen according to the fourth embodiment. [Figure 24] FIG. 11 is a schematic diagram showing a file list screen according to the fourth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] An information processing method, an information processing device, and a computer program according to embodiments of the present disclosure will be described below with reference to the drawings. Note that the present disclosure is not limited to these examples, but is defined by the claims, and is intended to include all modifications within the meaning and scope equivalent to the claims. Furthermore, at least some of the embodiments described below may be combined in any manner.

[0011] (Embodiment 1) 1 is a conceptual diagram showing the configuration of an information processing system according to embodiment 1. The information processing system according to this embodiment includes a cloud server (information processing device) 1, a large language model (LLM) server (hereinafter referred to as an LLM server 2) 2, and a user terminal 3. The cloud server 1, the LLM server 2, and the user terminal 3 are connected via a communication network N, and transmit and receive various data.

[0012] <Cloud Server 1> 2 is a block diagram showing an example of the configuration of the cloud server 1 according to the first embodiment. The cloud server 1 is a computer including a processing unit 11, a memory unit 12, a storage unit 13 for storing brain data, a working storage unit 14 for data analysis, and a communication unit 15. The cloud server 1 may be configured to perform distributed processing using multiple computers, or may be realized by multiple virtual machines provided in a single server, or may be partially configured by a quantum computer.

[0013] The storage unit 12 is a storage device such as a hard disk, an EEPROM (Electrically Erasable Programmable Read-Only Memory), a flash memory, etc. The storage unit 12 stores a computer program 121, a question-and-answer learning model 122, and various other programs and data.

[0014] The question-and-answer learning model 122 is, for example, a large language model (LLM), and can output questions related to the characteristics of the brain data based on the input brain data, metadata of the brain data, etc. By further inputting the question and answer into the question-and-answer learning model 122, additional questions can also be output.

[0015] The question-and-answer learning model 122, which is a large-scale language model, is a language generation model that, when a prompt consisting of an instruction sentence requesting some kind of answer and information related to the prompt are input, outputs a sentence corresponding to the input prompt and information. The information related to the prompt includes brain data, meta-information such as basic labels described below, and the like. The large-scale language model is a model based on a Transformer architecture, including GPT (Generative Pretrained Transformer, including GPT-1, GPT-2, GPT-3, GPT-4, etc.), BERT (Bidirectional Encoder Representations from Transformers), BART (Bidirectional and Auto-regressive Transformer), LLaMA (Large Language Model Meta AI), etc. The large-scale language model is pre-trained and is capable of handling various tasks without fine-tuning. Of course, the large-scale language model may also be fine-tuned by additional training.

[0016] The large-scale language model is realized by the processing unit 11 executing information processing in accordance with the computer program 121. The storage unit 12 stores data for realizing the large-scale language model. The large-scale language model may be configured with hardware. The large-scale language model may be realized using a quantum computer. Alternatively, the large-scale language model may be provided outside the cloud server 1, and the cloud server 1 may execute processing using the external large-scale language model. In other words, the question-and-answer learning model 122 may use one provided by an external API. Furthermore, the large-scale language model may be realized using multiple computers connected via a communication network, or may be realized using the cloud. Furthermore, the large-scale language model provided in the LLM server 2 described later may be used as the question-and-answer learning model 122.

[0017] The question-and-answer learning model 122 may be realized based on other machine learning methods such as neural networks and regression models, or statistical methods, etc. The function of the question-and-answer learning model 122 may also be realized by methods such as decision trees, random forests, linear regression, and Bayesian models.

[0018] The computer program 121 may be recorded in a computer-readable manner on the recording medium 10. The storage unit 12 stores the computer program 121 read from the recording medium 10 by a reading device (not shown). The recording medium 10 may be a semiconductor memory such as a flash memory, an optical disk, a magnetic disk, a magneto-optical disk, or the like. Alternatively, the various programs according to this embodiment may be downloaded from an external server (not shown) connected to the communication network N and stored in the storage unit 12.

[0019] The processing unit 11 has one or more arithmetic processing devices such as a CPU (Central Processing Unit) or an MPU (Micro-Processing Unit), and implements the information processing method of this embodiment by reading and executing a computer program 121 stored in the memory unit 12.

[0020] The large-capacity storage 13 and working storage 14 for storing brain data are storage devices such as hard disks, and have the same hardware configuration as the storage unit 12. The storage 13 stores account information for multiple users and brain data uploaded and saved by the users. The account information includes contact information such as a user ID, password, and email address, as well as service plan information. The service plan information indicates, for example, the storage capacity provided to the user and the services available to them. The cloud server 1 provides the storage capacity of the storage 13 to each of multiple users according to the account information. In other words, the user can use the storage 13 to save brain data up to the storage capacity. The brain data saved in the storage 13 includes brain data obtained from a living human brain. The brain data includes not only brain data of a living human brain obtained by non-invasive brain function measurement, but also brain data of a living human brain obtained directly by invasive measurement (invasive raw brain data). Needless to say, the brain data storage 13 and the working storage 14 may be configured with a plurality of storage devices.

[0021] The communication unit 15 includes a processing circuit, a communication circuit, and the like for performing processing related to communication, and transmits and receives information between the LLM server 2 and the user terminal 3.

[0022] <LLMサーバ2> The LLM server 2 is a server equipped with a general-purpose machine learning model 21 (see FIG. 4). The general-purpose machine learning model 21 is, for example, a large language model (LLM), and is a machine learning model that is trained so that when data to be analyzed is given and a prompt describing a data analysis method, a data output method, etc. is input, the machine learning model outputs a data analysis result according to the instructions of the prompt.

[0023] The general-purpose machine learning model 21 has the same configuration as the large-scale language model of the question-and-answer learning model 122. The general-purpose machine learning model 21 is a language generation model that, when a prompt consisting of an instruction sentence requesting an answer and brain data are input, outputs a sentence corresponding to the input prompt and brain data. The large-scale language model is a model based on a transformer architecture including GPT, BERT, BART, LLaMA, etc. The large-scale language model is pre-trained and is a language model capable of handling various tasks without fine-tuning. Of course, the large-scale language model may be fine-tuned by additional training. The general-purpose machine learning model 21 is not particularly limited in type as long as it is a model that can execute natural language processing that processes input brain data, words, and sentences, and outputs the required results. Furthermore, although a "general-purpose" machine learning model has been described because it can be used for various processes, it may also be a large-scale language model specialized for analyzing brain data.

[0024] The prompt input to the general-purpose machine learning model 21 is an instruction to the general-purpose machine learning model 21, and includes a character string describing the desired data processing content. For example, when brain data and a prompt requesting data analysis processing for the brain data are input to the general-purpose machine learning model 21, the general-purpose machine learning model 21 can analyze the brain data in response to the prompt and output the analysis results. The form of the analysis results is not limited, and may include documents written in natural language, summaries, images, graphs, lists, etc. Then, the LLM server 2 executes the analysis process of the brain data in response to at least a request from the cloud server 1, and transmits the analysis results of the brain data.

[0025] The data analysis process is an example of a process performed by the general-purpose machine learning model 21, and may perform graph generation process, program generation process, document generation process such as application form generation process, and process of outputting PDF or URL of related papers, etc., depending on the input brain data and prompts.

[0026] Furthermore, an example will be described in which data analysis of brain data is performed using an external LLM server 2, but the cloud server 1 may be configured to include a general-purpose machine learning model 21 and the processing unit 11 may perform data analysis.

[0027] <User terminal 3> The user terminal 3 is a communication device equipped with a terminal control unit, a terminal storage unit, a terminal communication unit, an operation unit, and a display unit. The user terminal 3 is a smartphone, a tablet terminal, a personal computer, etc. The basic configurations of the terminal control unit, the terminal storage unit, and the terminal communication unit are the same as those of the cloud server 1. The terminal storage unit communicates with the cloud server 1 and stores programs for performing basic processes such as uploading brain data, receiving and displaying the results of brain data analysis, for example, a program related to a browser. The operation unit is a keyboard, a mouse, etc., and accepts user operations and inputs. The display unit is a display device such as a liquid crystal display or an organic EL display, and displays information provided by the terminal control unit.

[0028] <Storage and processing of brain data> Fig. 3 is a flowchart showing the processing procedure for storing and labeling brain data, Fig. 4 is a conceptual diagram showing a method for storing brain data, and Fig. 5 is a conceptual diagram showing brain data stored in storage 13. Here, it is assumed that the user has an account on cloud server 1, has logged in to cloud server 1, and has established a session between cloud server 1 and user terminal 3.

[0029] The processing unit 11 of the cloud server 1 controls the GUI (Graphical User Interface) of the user terminal 3. The processing unit 11 receives a brain data saving instruction through the cloud server 13 interface (step S111). For example, the processing unit 11 transmits to the user terminal 3 information such as the storage capacity allocated to the user account, the directory structure, the total storage capacity, the used capacity, the unused capacity, operations on file data stored in the cloud server 1, and information related to the interface structure for accepting data upload operations. The user terminal 3 receives this information transmitted from the cloud server 1, displays the status of the storage 13 on the browser, and accepts operations such as data upload and download.

[0030] The user can operate the upload of brain data using the user terminal 3. For example, by dragging and dropping a file of brain data onto a screen showing the storage area of the storage 13 allocated to the user account, the user can instruct the saving of the brain data and upload the brain data to the cloud server 1.

[0031] The processing unit 11 of the cloud server 1 stores the brain data uploaded from the user terminal 3 in the storage 13 in association with the user ID (step S112). The processing unit 11 may be configured to display the progress of uploading and storing the brain data. For example, the processing unit 11 may display text such as "XX% uploaded." The processing unit 11 may also accept multiple uploads of brain data and perform the storing process of multiple pieces of brain data in parallel. Furthermore, since there are various file formats for brain data, the processing unit 11 may be configured to convert the file format of the uploaded brain data into a common file format, such as ".mat" or ".csv".

[0032] Next, the processing unit 11 calculates the noise level of the brain data stored in step S112, and stores a noise level data file indicating the calculated noise level in association with the brain data in the storage 13 (step S113). The noise level may be calculated using a noise evaluation algorithm called FOOOF, for example. For example, a brain data ID is assigned to the brain data, and the same brain data ID is assigned to the noise level data file. The name of the noise level data file may be such that the corresponding brain data can be identified, such as "XXX data uploaded on MM / DD / YYYY at hh:mm." Note that the method of associating brain data with a noise level data file is merely an example and is not particularly limited. For example, information indicating the degree of noise may be included as metadata in the brain data file.

[0033] Next, the processing unit 11 acquires basic labels such as the species of animal that is the detection source of the uploaded brain data, the type of brain data, the presence or absence of a disease in the animal that is the detection source of the brain data or the name of the disease, the data acquisition date, task details, etc. via the user terminal 3 (step S114). The basic labels (first metadata) are metadata that are generally common to all brain data. The basic labels are information obtained by asking the user standard questions. The animal species is information indicating the type of animal, such as human, monkey, mouse, rat, marmoset, etc. The type of brain data is information indicating the type of data, such as "single unit," "LFP," "ECoG," or "EEG."

[0034] All or part of the basic labels may also be acquired by the process of step S115. However, it is preferable that the metadata of basic information is acquired by standard questions based on a rule base.

[0035] Furthermore, in order to reduce the user's effort of inputting the same basic label information each time brain data is uploaded, the basic label information may be stored in association with the user's account, and the next time brain data is uploaded, the basic label information previously input may be presented as an input candidate or automatically input.

[0036] Next, the processing unit 11 acquires macro labels (second metadata) related to the brain data by asking questions about the uploaded brain data via the user terminal 3 (step S115). Macro labels are metadata of the brain data obtained through dialogue with the user. Macro labels cannot be obtained by standard questions, but are information obtained by asking non-standard questions to the user. The processing unit 11 acquires the metadata of the brain data by engaging in dialogue with the user using the question-and-answer learning model 122. For example, brain data is provided to the question-and-answer learning model 122, and a prompt is input to the question-and-answer learning model 122 requesting that a question be created to elicit metadata that indicates the characteristics of the brain data, thereby creating a question to elicit metadata of the brain data from the user. Note that the brain data and basic label metadata may be provided to the question-and-answer learning model 122, and a prompt may be input to cause the model to output a question for eliciting the metadata.

[0037] Next, the processing unit 11 acquires, via the user terminal 3, microlabels (third metadata) associated with each time point of the brain data, which is time-series data (step S116). This is metadata assigned to each EEG measurement value arranged in chronological order. The user can upload, for example, tabular data associating EEG measurement values with microlabels to the cloud server 1 using the user terminal 3. The processing unit 11 of the cloud server 1 acquires the microlabels by receiving and reading the uploaded tabular data.

[0038] Next, the processing unit 11 creates dictionary-format data of the brain data based on the brain data and the basic labels, macro labels, and micro labels acquired in steps S114 to S116, associates it with the brain data, and stores it in the storage 13 (step S117), and ends the brain data storage process.

[0039] Although basic labels, macro labels, micro labels, etc. are associated with the brain data through the above process, the user cannot recognize the metadata. Therefore, the processing unit 11 may be configured to read all or part of the basic labels and micro labels from the storage 13 and display them in a pop-up window when the cursor is placed on an uploaded brain data file or when other predetermined operations are performed.

[0040] As shown in Fig. 5, the dictionary-format data is data in which numbers (No.) corresponding to time points in a time series, time-series electroencephalogram measurement values, brain regions where the measurement values were obtained, basic labels, microlabels, and macrolabels are associated with each other. The dictionary-format data is, for example, data in a table format. For example, a brain data ID is assigned to the dictionary-format data, and the dictionary data is associated with the brain data.

[0041] <Data analysis and processing of brain data stored in Storage 13> Fig. 6 is a flowchart showing the processing procedure for data analysis using brain data stored in the storage 13, and Fig. 7 is a conceptual diagram showing a data analysis processing method using brain data stored in the storage 13. The processing unit 11 of the cloud server 1 reads out the brain data stored in the storage 13 and transfers it to the working storage 14 (step S131). Then, the processing unit 11 creates one or more prompt candidates for making a data processing request to the LLM server 2 based on the noise level data, basic labels, macro labels, micro labels, etc. (step S132). The prompt candidates may be created on a rule-based basis, or the brain data may be provided to the general-purpose machine learning model 21, which may output data indicating an analysis method and a data processing command method.

[0042] Examples of prompts for data processing are: Example 1: "Display the first 5 rows and 1,000 columns of the raw data in a table, then plot the data from electrode 1 as a waveform, then apply a 50 Hz notch filter to all the data, then apply a Fourier transform to each row, and plot the results from electrode 1." Example 2: "Use this data to build a machine learning algorithm that can interpret the data to identify the words the subject is thinking about."

[0043] While the example described above presents prompt candidates to be input to the general-purpose machine learning model 21, brain data may be fed to the general-purpose machine learning model 21, and information suggesting a method for analyzing the brain data may be output from the general-purpose machine learning model 21 and provided to the user. Examples of information suggesting an analysis method are as follows: Example 1: "This data is XX (LFP, ECoG, EEG) data recorded on X month X day from XX (brain region) of XX (animal species) (with XX disease) at a sampling frequency of X Hz." Example 2: "This data is X rows and X columns, with the rows representing time and the columns representing electrode numbers." Example 3: "Display the first 5 rows and 1,000 columns of this data in a table format and plot the first 3 rows on a graph." Example 4: "Evaluate the level of noise in this data." Example 5: "Plot the power spectrum of the first 3 channels on a graph and display the FOOOF results." Example 6: "Present the FOOOF results for all channels as a CSV file." Example 7: "Channel XX has a very high level of noise, so we recommend not performing analysis using these channels."

[0044] The processing unit 11 may also be configured to present the following analysis methods to the user for selection. Examples of analysis methods include downsampling, calculation of various statistics (mean, median, variance, standard deviation), ERP identification, wavelet transform, correlation calculation, cross-correlation calculation, causal inference, dimensionality reduction (PCA, t-SNE, UMAP, ICA), and various tests (t-test, ANOVA, etc.). A program suitable for the brain data to be analyzed may be selected and acquired from a server storing programs for analyzing brain data. Regarding the program selection, the processing unit 11 may also be configured to determine the optimal program by inputting the brain data and a prompt instructing program selection to the general-purpose machine learning model 21.

[0045] Next, the processing unit 11 transmits one or more prompt candidates created in step S132 to the user terminal 3, thereby providing the prompt candidates to the user (step S133).

[0046] Next, the processing unit 11 acquires the prompt determined by the user via the user terminal 3 (step S134). The user can select a desired prompt candidate from one or more prompt candidates, or create a prompt candidate by editing the prompt candidate. The user operates the user terminal 3 to send the determined prompt to the cloud server 1, and the cloud server 1 acquires the prompt candidate sent from the user terminal 3.

[0047] The processing unit 11 transmits the brain data stored in the working storage 14 and the prompt acquired in step S134 to the LLM server 2, and acquires the data processing results obtained by analysis based on the brain data and the prompt (step S135). The LLM server 2 inputs the brain data and the prompt transmitted from the cloud server 1 into the LLM model, thereby executing data analysis processing according to the contents of the prompt, and transmits the data processing results obtained by the analysis processing to the cloud server 1. The cloud server 1 receives the data processing results transmitted from the LLM server 2, thereby acquiring the data processing results.

[0048] Next, the processing unit 11 transmits the data processing results of the brain data to the user terminal 3 that has requested the data analysis (step S136), and ends the processing. The user terminal 3 receives and displays the received data processing results.

[0049] <Brain data analysis and transfer storage> 8 and 9 are flowcharts showing the processing procedures for uploading brain data to perform data analysis and data transfer / storage, FIG. 10 is a conceptual diagram showing a data analysis processing method for uploading brain data, and FIG. 11 is a conceptual diagram showing a method for transferring and storing brain data to be analyzed in storage 13.

[0050] The processing unit 11 of the cloud server 1 accepts the analysis process of the brain data through the user terminal 3 (step S151). For example, the user terminal 3 accepts the selection and upload operation of the brain data to be analyzed, and can operate the upload of the brain data using the user terminal 3.

[0051] The processing unit 11 of the cloud server 1 stores the brain data uploaded from the user terminal 3 in the working storage 14 in association with the user ID (step S152).

[0052] Next, the processing unit 11 calculates the noise level of the brain data stored in step S152 (step S153). Furthermore, the processing unit 11 acquires, via the user terminal 3, basic labels such as the species of the animal that is the detection source of the uploaded brain data, the type of brain data, and the presence or absence or name of a disease in the animal that is the detection source of the brain data (step S154). Furthermore, the processing unit 11 acquires, via the user terminal 3, a macro label (second metadata) related to the brain data by asking a question related to the uploaded brain data (step S155). Furthermore, the processing unit 11 acquires, via the user terminal 3, a micro label (third metadata) associated with each time point of the brain data, which is time-series data (step S156).

[0053] Next, the processing unit 11 provides prompt candidates by performing processing similar to steps S132 to S136, obtains the prompt candidates determined by the user, performs data analysis using the LLM model, and transmits the data processing results to the user terminal 3 (steps S157 to S161).

[0054] Next, the processing unit 11 accepts a decision as to whether or not to store the uploaded brain data as a data analysis target in the storage 13 (step S162). Note that the brain data may be stored without performing the determination process of step S162. If it is determined not to save in step S162 (step S162: NO), the processing unit 11 ends the process without saving the brain data, and the brain data in the working storage 14 is deleted.

[0055] If it is determined to save (step S162: YES), the processing unit 11 transfers the brain data stored in the working storage 14 as the data analysis target to the storage 13, associates it with the user account, and stores it in the storage 13 (step S163).

[0056] Next, the processing unit 11 stores a noise level data file indicating the noise level calculated in step S153 in association with the brain data in the storage 13 (step S164). Then, the processing unit 11 creates dictionary-format data of the brain data based on the brain data and the basic labels, macro labels, and micro labels acquired in steps S154 to S156, and stores the dictionary-format data in association with the brain data in the storage 13 (step S165), thereby completing the brain data storage process. The basic label may be stored as metadata in association with the brain data.

[0057] According to the information processing system configured as described above, a storage area for storing brain data is provided for a user's account, and the convenience of storing and analyzing brain data can be improved, taking into consideration users who are not experts in data processing.

[0058] In this embodiment, an example of analyzing brain data using the general-purpose machine learning model 21 has been described, but the memory unit 12 may be configured to store one or more data analysis programs, and the processing unit 11 may be configured to analyze brain data using the data analysis programs.

[0059] Furthermore, in this embodiment, an example has been described in which the general-purpose machine learning model 21 is mainly used to perform the analysis processing of brain data, but the general-purpose machine learning model 21 may also be configured to provide various functions related to brain data and brain research. By inputting brain data into the general-purpose machine learning model 21 and prompts that instruct the required processing content, processing conditions, data output method, etc., various processing results can be obtained. For example, the processing unit 11 may input a PDF file of a paper uploaded by a user or a URL indicating the location of the PDF file into the general-purpose machine learning model 21, and output an explanation of the paper. The processing unit 11 transmits data relating to the explanation of the paper to the user terminal 3. The processing unit 11 may also input the brain data into the general-purpose machine learning model 21 and output information such as the titles and locations of papers related to the brain data. The processing unit 11 transmits information on the related papers to the user terminal 3. Furthermore, the processing unit 11 may input paper data based on the brain data uploaded by the user to the general-purpose machine learning model 21 and output the peer review results of the paper. The processing unit 11 transmits data related to the commentary on the paper to the user terminal 3. Furthermore, the processing unit 11 may input the brain data into the general-purpose machine learning model 21 and cause the brain data to output draft data of various applications or papers. The processing unit 11 transmits the application data or the draft data of the paper to the user terminal 3. Furthermore, the processing unit 11 may input the brain data into the general-purpose machine learning model 21 and output research support data indicating hypotheses obtained from the brain data, additional experiments to be conducted, etc. The processing unit 11 transmits the research support data to the user terminal 3. Furthermore, the processing unit 11 may input the brain data into the general-purpose machine learning model 21 and output data that has been processed to make it easier to analyze the brain data. The processing unit 11 transmits research support data to the user terminal 3.

[0060] Furthermore, in this embodiment, the cloud server 1 has been described as one that mainly stores and analyzes brain data, but it may also be configured to store and analyze data obtained in other research fields and technology development fields besides brain data.

[0061] Furthermore, the more information related to basic labels is provided, or depending on the value of brain data such as brain-invasive raw data, the lower the usage fee for the cloud server 1 may be, or the range of use of the general-purpose machine learning model 21 may be expanded, thereby providing incentives to users.

[0062] (Appendix 1) 1. An information processing method for providing a storage area for storing brain data for a user's account, comprising: The uploaded brain data is associated with the user's account and stored in storage. obtaining metadata associated with the brain data; The acquired metadata is stored in association with the brain data. Information processing methods. (Appendix 2) reading the brain data stored in the storage and providing it to a machine learning model, and obtaining a data processing result by inputting a prompt requesting data processing; Providing the obtained data processing results to the user 1. The information processing method described in Appendix 1. (Appendix 3) providing the uploaded brain data to the machine learning model and obtaining a data processing result by inputting a prompt requesting data processing; providing the acquired data processing results to a user; The brain data to be processed is stored in the storage in association with a user account. 1. The information processing method described in Appendix 2. (Appendix 4) Calculating a noise level of the uploaded brain data; storing the calculated noise level in the storage in association with the brain data; providing the brain data to the machine learning model and obtaining the data processing results by inputting a prompt requesting data processing, the prompt including a character string related to the noise level associated with the brain data; 1. An information processing method according to claim 2 or 3. (Appendix 5) acquiring metadata including at least one of the species of the animal that is the detection source of the uploaded brain data, the type of the brain data, and the presence or absence of a disease or the name of the disease of the animal that is the detection source of the brain data; storing the acquired metadata in the storage in association with the brain data; providing the brain data to the machine learning model and obtaining the data processing results by inputting a prompt requesting data processing, the prompt including a character string related to the metadata associated with the brain data. 10. An information processing method according to any one of claims 2 to 4. (Appendix 6) acquiring first metadata including at least one of the species of the animal that is the detection source of the uploaded brain data, the type of the brain data, and the presence or absence of a disease or the name of the disease of the animal that is the detection source of the brain data; obtaining second metadata about the uploaded brain data by querying the uploaded brain data; acquiring third metadata associated with each time point of the brain data, which is time-series data; The acquired first to third metadata are associated with the brain data and stored in the storage. 6. An information processing method according to any one of claims 1 to 5.

[0063] (Embodiment 2) The cloud server 1 according to the second embodiment differs from the cloud server 1 according to the first embodiment mainly in the metadata assignment process. Since the other configurations of the cloud server 1 are the same as those of the cloud server 1 according to the first embodiment, the same reference numerals are used for the same parts and detailed description will be omitted.

[0064] <LLMサーバ2> In the first embodiment, an example was described in which the LLM server 2 is an off-premises large-scale language model, but the LLM server 2 according to the second embodiment will be described as an on-premises server. That is, the LLM server 2 can communicate with the cloud server 1, but the LLM server 2 is not connected to the public communication network N, and the user terminal 3 cannot communicate directly with the LLM server 2. Note that the LLM server 2 may be an on-premises server, as in the second embodiment.

[0065] <Storage of brain data and metadata acquisition processing> FIG. 12 is a flowchart showing the processing procedure for storing brain data and assigning metadata according to the second embodiment, and FIGS. 13 and 14 are schematic diagrams showing an example of a metadata input screen during uploading of brain data according to the second embodiment.

[0066] The processing unit 11 of the cloud server 1 receives an instruction to save the brain data through the GUI of the user terminal 3 (step S211). Then, in parallel with the process of storing the brain data uploaded from the user terminal 3 in the storage 13 in association with the user ID, the processing unit 11 of the cloud server 1 acquires metadata and stores the metadata in association with the brain data in the storage 13 (step S212). That is, since uploading large amounts of brain data takes time, the processing unit 11 starts the process of acquiring metadata related to the brain data while uploading the brain data.

[0067] The metadata acquired during uploading of brain data includes, for example, the animal species, the presence or absence of a brain disease or the name of the disease, the presence or absence of a gene mutation, the type of brain data, the manufacturer name of the recording device used to record the brain data, the sampling frequency, the number of recording channels, and the task details. The type of brain data includes, for example, scalp electroencephalogram (EEG), electrocorticogram (ECoG), local field electroencephalogram (LFP), spike data, etc.

[0068] For example, the processing unit 11 displays the animal species selection screen 41 shown in FIG. 13A via the user terminal 3 and acquires metadata indicating the animal species for which brain data was detected. The animal species selection screen 41 includes letters and numbers indicating what percentage of the brain data has been uploaded. The animal species selection screen 41 also includes, for example, a "human" icon, a "mouse" icon, a "rat" icon, a "monkey" icon, and an "other" icon for selecting the animal species. The user can select the animal species for which brain data was detected by clicking these icons.

[0069] Furthermore, the processing unit 11 displays a brain disease presence / absence selection screen 42 shown in Fig. 13B and acquires metadata indicating the presence or absence of a brain disease. The brain disease presence / absence selection screen 42 includes radio buttons for selecting the presence or absence of a brain disease. The user can select the presence or absence of a brain disease by operating the radio buttons.

[0070] Furthermore, the processing unit 11 displays a genetic mutation presence / absence selection screen 43 shown in Fig. 13C and acquires metadata indicating the presence or absence of a genetic mutation. The genetic mutation presence / absence selection screen 43 includes radio buttons for selecting the presence or absence of a genetic mutation. The user can select the presence or absence of a genetic mutation by operating the radio buttons.

[0071] Furthermore, the processing unit 11 displays a data type selection screen 44 shown in Fig. 14A and acquires metadata indicating the type of brain data. The data type selection screen 44 includes, for example, a "scalp electroencephalogram (EEG)" icon, an "electrocorticogram (ECoG)" icon, a "local field electroencephalogram (LEP)" icon, a "spike data" icon, and an "other" icon for selecting the type of brain data. The user can select the type of brain data by clicking these icons.

[0072] 14B, and acquires metadata indicating the name of the recording device manufacturer used to detect the brain data, the sampling frequency of the brain data, the number of recording channels, and task content related to the analysis of the brain data. The recording details / task content input screen 45 includes, for example, a text input field for inputting the name of the recording device manufacturer, a text input field for inputting the sampling frequency of the brain data, a text input field for inputting the number of recording channels, and a text input field for inputting task content related to the analysis of the brain data.

[0073] Although the methods of accepting metadata have been exemplified by clicking an icon, selecting a radio button, and entering text into a text field, the methods of accepting input of metadata are not limited to these.

[0074] Next, even after the storage process of the uploaded brain data is completed, the processing unit 11 acquires and edits the metadata and stores the metadata in association with the brain data (step S213). For example, the processing unit 11 can additionally execute a metadata acquisition process, a metadata editing process, etc. on a brain data analysis screen for analyzing the brain data.

[0075] <Brain data analysis screen and workspace> FIG. 15 is a schematic diagram illustrating an example of the brain data analysis screen 5 according to the second embodiment. The brain data analysis screen 5 includes a chat interface 51 that displays prompts to be input into a machine learning model and responses output from the machine learning model, and a workspace 52 that displays acquired and unacquired metadata and other data. The workspace 52 includes a data summary tab 52a, an analysis data tab 52b, and a data content tab 52c. FIG. 15 illustrates a state in which the data summary tab 52a is selected. When the data summary tab 52a is selected, the processing unit 11 displays the metadata of the brain data selected for analysis. Here, the names of entered metadata and unentered metadata are displayed. The contents and names of the metadata include, for example, an overview of the experiment and data, the presence or absence of a gene mutation, the presence or absence of a disease, the type of brain data, the manufacturer of the recording device, the sampling frequency, the number of channels, the animal species, the individual's name, the brain region, and the recording date. The user can freely edit the metadata displayed when the data summary tab 52a is selected. The processing unit 11 accepts input and editing of metadata by the user, and stores the newly input metadata and the changed metadata in association with the brain data.

[0076] Fig. 16 is a schematic diagram showing a state in which the analysis data tab 52b of the workspace 52 is selected. When the analysis data tab 52b of the workspace 52 is selected by a user operation, the processing unit 11 displays a list of files related to the analysis processing of the brain data, such as the file of the brain data to be analyzed and the file generated by the data processing, as shown in Fig. 16. Furthermore, when one file is selected from the files displayed when the analysis data tab 52b is operated, the processing unit 11 automatically selects the data content tab 52c and displays the data content of the selected one file.

[0077] Fig. 17 is a schematic diagram showing a state in which the data content tab 52c of the workspace 52 is selected. When the data content tab 52c of the workspace 52 is selected by a user operation, the processing unit 11 displays the specific contents of the data of the file selected as the target of data processing, as shown in Fig. 17.

[0078] The workspace 52 may be configured to include an image tab. When the image tab of the workspace 52 is selected by a user operation, the processing unit 11 displays a list of images such as graphs obtained by the analysis processing of the brain data. When an image is selected by a user operation, the processing unit 11 causes the chat interface 51 to display the prompt input or the response sentence output when generating the selected image. For example, the processing unit 11 scrolls the chat interface 51 to the position of the prompt and the response sentence.

[0079] In the second embodiment, metadata of the brain data can be added using the upload time of the brain data.

[0080] In the second embodiment, an example in which a user basically inputs metadata has been described, but the processing unit 11 may be configured to automatically identify and assign some metadata using a learning model for identifying metadata. The learning model for identifying metadata is, for example, a neural network model that has been trained by machine learning so that, when brain data is input, the model outputs the animal species of the detection source of the brain data, the type of brain data, the sampling frequency, the number of channels, etc. The processing unit 11 may also be configured to identify metadata of brain data based on information input by a user during interactive interaction with the user, and store the identified metadata in association with the brain data. While the method for identifying metadata is not particularly limited, for example, it is preferable to identify the metadata of brain data using a language processing model such as LLM or BART (Bidirectional Encoder Representations from Transformers). Specifically, the metadata can be extracted by inputting, into the language processing model, an instruction statement that includes the information input by the user and the name of the metadata and that instructs the language processing model to extract the metadata from the information.

[0081] (Embodiment 3) The cloud server 1 according to the third embodiment differs from the cloud server 1 according to the first and second embodiments mainly in the method of analyzing and processing brain data. The other configurations of the cloud server 1 are the same as those of the cloud server 1 according to the first and second embodiments, so the same reference numerals are used for the same parts and detailed descriptions are omitted.

[0082] <Data analysis and processing of brain data stored in Storage 13> FIG. 18 is a flowchart showing the processing procedure of data analysis according to the third embodiment using brain data stored in the storage 13, and FIG. 19 is a conceptual diagram showing the data analysis processing method according to the third embodiment using brain data stored in the storage 13.

[0083] The processing unit 11 of the cloud server 1 reads out the brain data stored in the storage 13 and transfers it to the working storage 14 (step S331). Then, the processing unit 11 creates one or more prompt candidates for making a data processing request or a data processing program request to the LLM server 2 based on the metadata (step S332). The data processing program request means a request to generate and transmit a data processing program required for executing the brain data analysis process on the cloud server 1 side.

[0084] Next, the processing unit 11 provides the prompt candidates to the user by transmitting one or more prompt candidates created in step S332 to the user terminal 3 (step S333). The processing unit 11 acquires the prompt determined by the user via the user terminal 3 (step S334).

[0085] 7, the processing unit 11 transmits the brain data stored in the working storage 14 and the prompt for the data processing request acquired in step S334 to the LLM server 2, and acquires the data processing result obtained by analysis based on the brain data and the prompt (step S335). Alternatively, as shown in FIG. 19, the processing unit 11 transmits the prompt for the data processing program request acquired in step S334 to the LLM server 2, acquires the data processing program obtained by processing based on the prompt, and executes the acquired data processing program to acquire the data processing result (step S335).

[0086] Then, the processing unit 11 transmits the data processing result of the brain data to the user terminal 3 that has requested the data analysis (step S336). Then, the processing unit 11 adds the data obtained by the data processing of the brain data to the analysis data tab 52b of the workspace 52 (step S337), and ends the processing. The user terminal 3 receives and displays the received data processing result.

[0087] <Brain data analysis and transfer storage> FIG. 20 is a flowchart showing the processing procedure for data analysis and data transfer / storage according to the third embodiment, which is performed by uploading brain data; FIG. 21 is a conceptual diagram showing a data analysis processing method according to the third embodiment, which is performed by uploading brain data; and FIG. 22 is a conceptual diagram showing a method for transferring and storing brain data to be analyzed in the storage 13 according to the third embodiment.

[0088] The processing unit 11 of the cloud server 1 accepts the analysis processing of the brain data through the user terminal 3 (step S351).

[0089] Next, as in embodiment 2, the processing unit 11 acquires metadata and stores it in the storage 13 in association with the brain data uploaded from the user terminal 3 in association with the user ID (step S352) in parallel with the process of storing the brain data in the storage 13.

[0090] Next, the processing unit 11 provides prompt candidates by performing the same processing as in steps S332 to S336, acquires the prompt candidates determined by the user, performs data analysis using the LLM model, and transmits the data processing results to the user terminal 3 (steps S353 to S357). The processing unit 11 also adds data obtained by data processing of the brain data to the analysis data tab 52b of the workspace 52 (step S358). When requesting the LLM server 2 to execute data processing, the processing unit 11 provides the analysis results of the brain data in the flow shown in FIGS. 10 and 11, as in the first embodiment. When requesting the LLM server 2 to execute a data processing program, the processing unit 11 provides the analysis results of the brain data in the flow shown in FIGS. 21 and 22.

[0091] Next, the processing unit 11 accepts a request as to whether or not to store the uploaded brain data as a data analysis target in the storage 13 (step S359). If it is determined in step S362 that the brain data should not be saved (step S359: NO), the processing unit 11 ends the process without saving the brain data. The brain data in the working storage 14 is deleted.

[0092] If it is determined to store the data (step S359: YES), the processing unit 11 transfers the brain data stored in the working storage 14 as the data analysis target to the storage 13 as shown in FIG. 22, and stores the data in the storage 13 in association with the user account (step S360).

[0093] <System Instructions> The processing unit 11 according to the third embodiment is configured to provide the general-purpose machine learning model 21 of the LLM server 2 with system instructions describing operational guidelines including a procedure for executing data processing of brain data or a procedure for creating a data processing program for executing data processing, and then execute data processing related to the analysis of the brain data or generate a data processing program.

[0094] The system instructions include, for example, the following directives: First instruction: The first step in brain data processing should be data preprocessing, the second step feature extraction, the third step statistical analysis, and the fourth step consideration of the data processing results.

[0095] Second instruction: Before executing data processing of brain data, the brain data file to be analyzed, the analysis method, and parameters should be presented to the user for confirmation.

[0096] Third instruction: When data processing of brain data is performed, the data resulting from the data processing should be plotted on a graph.

[0097] Fourth directive: If the user does not specify a file to analyze, data processing should be performed on the most recently generated file. By providing system instructions to the general-purpose machine learning model 21 of the LLM server 2, it is possible to interact with the user and process data in accordance with the above-mentioned instructions.

[0098] Fifth instruction: When analyzing brain data, not only generate but also automatically execute data processing program code.

[0099] Sixth instruction: After responding to a user's instruction, always suggest multiple "next steps."

[0100] Seventh instruction: If a file is generated as a result of the analysis, the generated file is automatically added to the "Analysis Data" tab (Analysis Data tab 52b) of the "AI Workspace" (Workspace 52).

[0101] <Action and effect> By providing the above-mentioned system instructions to the LLM server 2 in advance, it is possible to more accurately analyze the brain data.

[0102] In the third embodiment configured as described above, a data processing program required for analyzing brain data can be created in the LLM server 2, and the execution processing of the brain data can be executed in the cloud server 1. Therefore, brain data can be analyzed regardless of limitations on the data to be analyzed, limitations on the data calculation time, limitations on the calculation load, etc. imposed by the LLM server 2.

[0103] The cloud server 1 can process the brain data in accordance with the first instruction statement of the system instruction by performing data preprocessing on the brain data, executing feature extraction processing and statistical analysis processing on the preprocessed brain data, and then considering the results of the data processing. Furthermore, the cloud server 1 can execute the analysis of the brain data step by step while confirming the user's intention, that is, while confirming the file of the brain data to be analyzed, the analysis method, and the contents of the parameters, by using the second instruction sentence of the system instruction. Furthermore, the cloud server 1 can output the analysis results as a graph using the third directive of the system instruction. Furthermore, when executing the analysis process, the cloud server 1 can set the most recently generated data as the analysis target and execute the data process unless instructed by the user, by the fourth instruction sentence of the system instruction. For example, when brain data to be analyzed is selected and a first process is executed on the brain data, if the processing unit 11 instructs to execute an additional second process, the processing unit 11 can execute the second process on the data on which the first process was executed, instead of the brain data.

[0104] When using an off-premises LLM server 2 outside the cloud server 1, if no data is input for a predetermined time, the session of the LLM server 2 may be closed. Therefore, the processing unit 11 may be configured to periodically input predetermined data into the LLM server 2 so as not to close the session even if no instruction is received from the user for a predetermined time or longer.

[0105] Furthermore, although an example in which one type of LLM server 2 is used has been described, if multiple LLM servers 2 can be used, the processing unit 11 may be configured to select an LLM server 2 depending on the situation.

[0106] Furthermore, when specific programming code is sent to the cloud server 1 by user operation, the processing unit 11 may be configured to receive the programming code sent from the user terminal 3, execute the received programming code, or have the LLM server 2 execute it. In addition, when a URL containing programming code information is sent to the cloud server 1 by user operation, the processing unit 11 may be configured to receive the URL sent from the user terminal 3, access the received URL to obtain the programming code, and execute the obtained programming code or have the LLM server 2 execute it.

[0107] Furthermore, when a screenshot image of a graph is provided by a user operation, the processing unit 11 may be configured to create a graph like the screenshot image based on the brain data. For example, in the case of a multimodal LLM server 2, by inputting the brain data metadata, brain data, and screenshot image, it is possible to generate a graph obtained by analyzing the brain data.

[0108] Furthermore, the cloud server 1 may be configured to save the workflow of data analysis performed through a series of interactions with the user as a template, and to call up and use the saved template repeatedly from the next time onwards in response to a user request.

[0109] (Embodiment 4) The cloud server 1 according to the fourth embodiment differs from the cloud server 1 according to the first to third embodiments in the processing method for adding metadata to folders and displaying folder lists and file lists. The other configurations of the cloud server 1 are the same as those of the cloud server 1 according to the first to third embodiments, so the same reference numerals are used for the same parts and detailed description will be omitted.

[0110] <Folder list> In the fourth embodiment, the processing unit 11 of the cloud server 1 associates the metadata with a plurality of files by adding the metadata to a folder storing the plurality of files. A user does not need to add metadata to each of the plurality of files, but can associate the metadata with the plurality of files by storing a plurality of files with similar attributes in one folder and adding the metadata to the folder. It is also possible to configure the system so that metadata common to all files stored in a folder is associated with the folder, and metadata that differs between individual files is associated with the individual files.

[0111] Fig. 23 is a schematic diagram showing a folder list screen 6 according to embodiment 4. When the processing unit 11 receives an instruction to display a folder list, it displays the folder list screen 6 showing a list of folders stored in the storage 13, as shown in Fig. 23.

[0112] The folder list screen 6 includes a list 61 of the names of multiple folders, and has an "incomplete tag" label 63 indicating that metadata association has not been completed for each of the multiple folders. The folder list screen 6 also has an "AI auto tagging" button (auto tagging button) 62 that instructs automatic assignment of metadata using a learning model for one or more folders for which metadata association has not been completed.

[0113] When the user operates the "AI-based automatic tagging" button 62, the processing unit 11 identifies metadata based on the brain data of the files stored in the folder. For example, the processing unit 11 identifies the metadata by inputting part of the brain data into a learning model for identifying metadata. Then, the processing unit 11 displays the file list screen 7. The learning model for identifying metadata is, for example, a neural network model that has been trained to output, when brain data is input, the animal species from which the brain data was detected, the type of brain data, the sampling frequency, the number of channels, etc.

[0114] The folder list screen 6 also has a filter button 64 for filtering the folders to be displayed based on metadata. The filter button 64 includes a plurality of buttons for specifying metadata such as the recording device, animal species, recording method, frequency, and disease model. The user can specify specific metadata by operating the filter button 64. The processing unit 11 extracts folders stored in the storage 13 that have been assigned the selected specific metadata, and displays a list of the names of the extracted folders.

[0115] <File list> 24 is a schematic diagram showing a file list screen 7 according to embodiment 4. The file list screen 7 includes a file list display area 71 that displays a list of files stored in a folder that has been automatically tagged with metadata, and a sidebar 72 that displays metadata associated with the folder and each file.

[0116] The sidebar 72 displays metadata automatically identified by the learning model. It also displays the names of metadata that have not yet been acquired. The processing unit 11 allows the user to approve or not approve the automatically assigned metadata, and if it is inappropriate, accepts editing of the metadata. The user can approve the metadata content by manipulating the metadata displayed in the sidebar 72. For example, the processing unit 11 displays an approve button for approving the automatically assigned metadata and a reject button for rejecting it, and accepts the user's button operation. It is also possible to perform operations such as changing the automatically assigned metadata and inputting additional metadata. The processing unit 11 stores the approved or edited metadata in association with the brain data.

[0117] The file list display area 71 also has an analysis button 73 for each of a plurality of files. By operating the analysis button 73, the user executes data processing for the file corresponding to the operated analysis button 73. Specifically, the file corresponding to the operated analysis button 73 is set as an analysis target, and the brain data analysis screen 5 as shown in FIG. 15 is displayed. The user can instruct the analysis processing of the brain data through a question and answer session via the chat interface 51 on the brain data analysis screen 5.

[0118] In the fourth embodiment, metadata can be assigned on a folder-by-folder basis. Also, metadata can be assigned automatically. [Explanation of symbols]

[0119] 1: Cloud server 2: LLM server 3: User terminal 10: Recording media 11: Processing section 12: Storage section 13: Storage 14: Working storage 15: Communications Department 21: General-purpose machine learning model 121: Computer Programs 122: Question-and-answer learning model N: communication network

Claims

1. 1. An information processing method for providing a storage area for storing brain data for a user's account, comprising: The uploaded brain data is associated with the user's account and stored in storage. obtaining metadata associated with the brain data; The acquired metadata is stored in association with the brain data. Information processing methods.

2. reading the brain data stored in the storage and providing it to a machine learning model, and obtaining a data processing result by inputting a prompt requesting data processing; Providing the obtained data processing results to the user The information processing method according to claim 1 .

3. and reading out the metadata of the brain data stored in the storage and providing it to a machine learning model, and acquiring the data processing program by inputting a prompt requesting a data processing program for processing the brain data; Execute the acquired data processing program; Provide the user with the data processing results obtained by executing the data processing program. The information processing method according to claim 1 .

4. providing the uploaded brain data to the machine learning model and obtaining a data processing result by inputting a prompt requesting data processing; providing the acquired data processing results to a user; The brain data to be processed is stored in the storage in association with a user account. The information processing method according to claim 2 .

5. providing metadata associated with the uploaded brain data to a machine learning model and obtaining a data processing program for processing the brain data by entering a prompt requesting the data processing program; executes the acquired data processing program to process the brain data; Execute the data processing program; The brain data to be processed is stored in the storage in association with a user account. The information processing method according to claim 2 .

6. the machine learning model is a large-scale language model; A system instruction describing an operational guideline including a procedure for executing data processing of the brain data or a procedure for creating a data processing program for executing the data processing is given to the machine learning model, and then the data processing is executed or a data processing program is generated. The information processing method according to any one of claims 2 to 5.

7. The system instructions include: The data processing method includes instructions to perform data preprocessing as the first step, feature extraction processing as the second step, statistical analysis processing as the third step, and consideration of the data processing results as the fourth step. The information processing method according to claim 6.

8. The system instructions include: and an instruction to the effect that the file of the brain data to be analyzed, the analysis method, and parameters should be presented and confirmed before executing data processing of the brain data. The information processing method according to claim 6.

9. The system instructions include: When the data processing of the brain data is performed, the data resulting from the data processing should be plotted on a graph. The information processing method according to claim 6.

10. The system instructions include: Contains instructions to execute data processing on the most recently generated file if the user does not specify which file to analyze. The information processing method according to claim 6.

11. The system instructions include: Contains instructions to always suggest multiple "next steps" after responding to a user's instructions. The information processing method according to claim 6.

12. During uploading of the brain data, a process of acquiring metadata related to the brain data is initiated. The information processing method according to any one of claims 1 to 5.

13. After the upload of the brain data is completed, the process of acquiring metadata related to the brain data is continued. The information processing method according to claim 12.

14. The metadata acquired during uploading of the brain data includes at least the animal species, the presence or absence of a gene mutation, the presence or absence or name of a disease in the animal that is the source of the brain data, the type of the brain data, the manufacturer name of the recording device used to record the brain data, the sampling frequency, the number of recording channels, and the task content. The information processing method according to claim 12.

15. A brain data analysis screen is displayed, the brain data analysis screen including a chat interface that displays prompts to be input to the machine learning model and responses output from the machine learning model, and a workspace that displays the acquired and unacquired metadata. The information processing method according to any one of claims 2 to 5.

16. the workspace includes a plurality of workspace tabs; When a first workspace tab is selected, displaying the acquired and unacquired metadata; When the second workspace tab is selected, it displays the data processing target and the data processing result file. When the third workspace tab is selected, the data content of the data processing target or the data processing result is displayed. The information processing method according to claim 15.

17. If the fourth workspace is selected, it displays the image of the data processing results.

17. The information processing method according to claim 16.

18. When data processing for the brain data is completed, the file of the data processing results is added to the page of the second workspace tab.

17. The information processing method according to claim 16.

19. By assigning metadata to a folder storing a plurality of files, the metadata is associated with the plurality of files. The information processing method according to any one of claims 1 to 5.

20. A folder list screen is displayed, which lists each of a plurality of folders and has a label indicating whether or not metadata association has been completed for each of the plurality of folders.

20. The information processing method according to claim 19.

21. the folder list screen displays an automatic tagging button for automatically associating metadata with the folder for which metadata association has not been completed; When the automatic tagging button is operated, metadata is identified based on brain data of the files stored in the folder; Accept approval or edit of identified metadata; Associating the approved or edited metadata with the folder 21. The information processing method according to claim 20.

22. A file list display area displays a list of multiple brain data files stored in the storage, and a sidebar displays metadata associated with each file. The information processing method according to any one of claims 1 to 5.

23. the file list display area has an analysis button for each of a plurality of files; When the analysis button is operated, data processing of the file corresponding to the analysis button is executed.

23. The information processing method according to claim 22.

24. An information processing device that provides a storage area for storing brain data for a user's account, Storage that associates the uploaded brain data with the user's account and stores it; a communication unit for receiving metadata related to the brain data; a processing unit that associates the metadata received by the communication unit with the brain data and stores the same; An information processing device comprising:

25. 1. An information processing method for providing a storage area for storing brain data for a user's account, comprising: The uploaded brain data is associated with the user's account and stored in storage. obtaining metadata associated with the brain data; The acquired metadata is stored in association with the brain data. A computer program that causes a computer to execute a process.

Images