A data processing method and system for laboratory quality management based on artificial intelligence

Abstract

The present invention relates to the technical field of laboratory quality management, and specifically provides a data processing method and system for laboratory quality management based on artificial intelligence. The method first collects laboratory data, preprocesses the data to generate a unique hash value, and uploads it to the blockchain. The original data is stored in the IPFS distributed storage system. Then, the laboratory data is analyzed through AI algorithms, and the analysis results are fed back to the smart contract to trigger corresponding operations. Through the smart contract on the blockchain, the system can monitor the data in real time and automatically execute quality control operations according to preset rules. In addition, the data sharing mechanism allows cross-institutional data access, and the access rights of different roles are controlled through the blockchain permissions to ensure the privacy of the data. The present invention achieves the beneficial technical effects of high efficiency, security, and traceability in the process of laboratory quality management through intelligent data processing, transparent auditing, and decentralized management of the blockchain.

Description

Technical Field

[0001] The present invention relates to the technical field of laboratory quality management, and particularly relates to a data processing method and system for laboratory quality management based on artificial intelligence. Background Art

[0002] With the rapid development of science and technology, laboratories, as important bases for scientific research and technological innovation, the level of their quality management is directly related to the accuracy and reliability of scientific research results. However, traditional laboratory quality management methods often have problems such as low efficiency, many human errors, and difficulty in dealing with large amounts of data. Therefore, the research on data processing methods and systems for laboratory quality management based on artificial intelligence has emerged, aiming to improve the efficiency and accuracy of laboratory quality management by introducing artificial intelligence technology;

[0003] Searching the prior art, it is found that: a Chinese invention patent with the publication number CN109800591B discloses a data processing method and system for medical laboratory quality management. In this invention, laboratory-related data of each laboratory is stored in the database of the server, and at the same time, the terminals of each laboratory and the terminals of laboratory personnel can communicate with the server, realizing the integrated storage of laboratory-related data of each laboratory, which is convenient for centralized management. Furthermore, through the centralized management of data, it is beneficial to ensure the improvement of data quality;

[0004] With the expansion of laboratory scale and the increase in scientific research activities, the amount of generated data has increased sharply. Traditional quality management methods are difficult to effectively process and analyze these laboratory data, resulting in low management efficiency. And because the security level of test data in the laboratory is relatively high, there is a risk of data leakage during storage and transmission. Summary of the Invention

[0005] The purpose of the present invention is to address the problems in the background art and propose a data processing method and system for laboratory quality management based on artificial intelligence.

[0006] The technical solution of the present invention: A data processing method for laboratory quality management based on artificial intelligence includes the following implementation steps:

[0007] S1. The laboratory management system collects laboratory data. For data that cannot be automatically obtained, the user manually inputs the data through the laboratory management system interface;

[0008] S2. The laboratory management system aligns data from different sources through timestamps. The laboratory data is stored in a decentralized data storage system. The SHA-256 algorithm is used to generate a unique hash value for each piece of collected laboratory data. The hash value of the data and related metadata are stored on the laboratory data blockchain. The laboratory data blockchain automatically adds an accurate timestamp for each on-chain operation.

[0009] S3. The laboratory management system uses the Z-score standardization method to process laboratory data and transmits the laboratory data to the AI ​​analysis platform that has passed the data analysis qualification check. The AI ​​model in the AI ​​analysis platform extracts features, outputs analysis results, and inputs the analysis results into the smart contract in the laboratory data blockchain;

[0010] S4. Set up smart contracts on the laboratory data blockchain, define the quality control rules for laboratory data, and monitor laboratory data in real time. After receiving the analysis results from the AI ​​analysis platform, the smart contract automatically performs corresponding operations according to the rules, and automatically approves the test data that needs to be approved and verified according to the predefined approval rules, and then uploads each approval result to the laboratory data blockchain.

[0011] Preferably, the decentralized data storage system adopts the IPFS distributed storage system. After the data is stored in the IPFS distributed storage system, the system generates a unique IPFS hash.

[0012] Preferably, the laboratory data includes one or more of personnel data in the laboratory, various equipment data, sensor data, and manually recorded experimental conditions or experimental results.

[0013] Preferably, the specific implementation steps of S4 are as follows:

[0014] S4-1. Perform data verification in the smart contract to check whether the data meets the standards defined in the contract. The smart contract automatically executes the predetermined operation based on the verification results. When abnormal data is detected, the system automatically takes measures according to the smart contract rules and records all trigger conditions and notifications on the laboratory data blockchain.

[0015] S4-2. Smart contracts define approval process rules. When the set conditions are triggered, approval tasks are automatically generated and assigned to management personnel. The approval tasks include but are not limited to checking abnormal situations and confirming data accuracy.

[0016] S4-3. Notify the assigned personnel of the generated approval task and ask them to approve it. The approval request is generated through a smart contract, including the data of the triggering conditions and the recommended operations. The approver reviews the approval request and makes a decision. The approval result is recorded and uploaded to the chain through the smart contract.

[0017] S4-4. Record the approval operation including the approver, approval time, and approval result on the laboratory data blockchain to generate an approval record, ensuring the transparency and compliance of the approval process.

[0018] Preferably, the data analysis qualification verification process is as follows:

[0019] S5-1. The AI analysis platform sends a data analysis request Query to the laboratory management system;

[0020] S5-2. When the laboratory management system receives the data analysis request Query, it generates a random number Rt and sends it to the AI analysis platform;

[0021] S5-3. The AI analysis platform generates a random number Rr and calculates the verification information I = H(ID||Rr||Rt), and then transmits the random number Rr and the verification information I to the laboratory management system;

[0022] Wherein, ID is the number of the AI analysis platform, || is the string concatenation symbol, and H is the hash function;

[0023] S5-4. The laboratory management system performs operations on the authorized AI analysis platform numbers ID in the laboratory internal database in a traversal manner to determine whether there is an ID i (1≦i≦n) such that the equation H(ID i ||Rr||Rt) = I holds. If it exists, it means that the AI analysis platform has obtained authorization to analyze the laboratory data, and then the laboratory management system outputs the laboratory data to the AI model in the AI analysis platform.

[0024] Preferably, the AI model in the AI analysis platform performs feature extraction. The extracted features include: using the constructed time series feature extraction model, through the LSTM network, extracting time series features from the device operation data, using the Fourier transform to convert the laboratory data from the time domain to the frequency domain and extracting potential periodic features, using the constructed convolutional neural network to extract image features from the image data generated in the experiment, and using the constructed text feature extraction model, through natural language processing technology, using the BERT+BiLSTM+self attention+CRF model to perform keyword extraction and topic model analysis on the unstructured text data.

[0025] Preferably, the access method for laboratory data is as follows:

[0026] S7-1. The laboratory manager classifies the collected laboratory data through the laboratory management system and establishes an association relationship between the data type and the permission level: and

[0027] S7-2. Attach attributes to the above-classified laboratory data to construct an attribute set Ω:

[0028]

[0029] Among them, the attribute Att 1 represents public data, the attribute Att 2 represents restricted data, and the attribute Att 3 represents sensitive data;

[0030] S7-3. Construct the association relationship between attributes and the hash values of laboratory data

[0031] S7-4. When a user or institution needs to submit an access request and specify the type of data required for access, the smart contract on the laboratory data blockchain approves the data access request according to the predefined authorization rules, checks the identity, role permissions, and data attributes of the access party, and confirms whether the type of data to be accessed meets the authorization conditions based on the association relationship ;

[0032] S7-5. If the data access request of the access party is approved, the nodes on the laboratory data blockchain query the records on the laboratory data blockchain according to the request, obtain the IPFS hash of the laboratory data corresponding to the laboratory data hash value, and use the IPFS hash retrieved from the laboratory data blockchain to access the actual laboratory data in the IPFS distributed storage system.

[0033] The technical solution proposed by the present invention: A data processing system for laboratory quality management based on artificial intelligence, which is applicable to the data processing method for laboratory quality management based on artificial intelligence proposed above, and is characterized by including:

[0034] A laboratory management system for collecting and preprocessing laboratory data;

[0035] An AI analysis platform, with the operation carrier being an AI algorithm box, which contains a series of AI models for analyzing various types of laboratory data;

[0036] The laboratory data blockchain generates a unique hash value for laboratory data through a hashing algorithm, stores it in the laboratory data blockchain, generates an accurate timestamp for each piece of data, records all operation processes, forms a traceable audit record, and the smart contract in the laboratory data blockchain can automatically execute operations according to predefined rules, and can automatically trigger alarms or start the automatic adjustment and maintenance processes of equipment;

[0037] The IPFS distributed storage system is used to store laboratory data.

[0038] Compared with the prior art, the above technical solutions of the present invention have the following beneficial technical effects:

[0039] The present invention designs a data processing method and system for laboratory quality management based on artificial intelligence. This method combines artificial intelligence and blockchain technology. First, it collects laboratory data in real time, preprocesses the data to generate a unique hash value, and uploads it to the blockchain. The original data is stored in the IPFS distributed storage system to ensure the integrity and immutability of the data. Then, it analyzes and predicts the laboratory data through AI algorithms, and feeds the analysis results back to the smart contract to trigger corresponding operations. Through the smart contract on the blockchain, the system can monitor the data in real time and automatically execute quality control operations according to preset rules. In addition, the data sharing mechanism allows cross-institutional data access, and controls the access rights of different roles through the blockchain permissions to ensure the security and privacy of the data;

[0040] Through intelligent data processing, transparent auditing and decentralized management of the blockchain, the present invention achieves the beneficial technical effects of high efficiency, security and traceability in the process of laboratory quality management. Description of the Drawings

[0041] Figure 1 It is an architecture diagram of a data processing system for laboratory quality management based on artificial intelligence proposed by the present invention;

[0042] Figure 2 It is a block structure diagram of the laboratory data blockchain;

[0043] Figure 3 It is a network structure diagram of the time series feature extraction model;

[0044] Figure 4 It is a network structure diagram of the text feature extraction model. Detailed Embodiments

[0045] Example 1, as Figure 1As shown in the figure, a data processing system for laboratory quality management based on artificial intelligence proposed by the present invention includes: a laboratory management system, an AI analysis platform, a laboratory data blockchain, and an IPFS distributed storage system;

[0046] The laboratory management system is used to collect various types of data in the laboratory, including but not limited to various types of equipment data, sensor data, and text data;

[0047] The operation carrier of the AI analysis platform is an AI algorithm box, which is an intelligent edge application server, built-in with domestic high-performance AI computing power chips, with high-performance decoding and reasoning capabilities. Its core lies in using deep learning technology to customize a series of security models to analyze various types of laboratory data;

[0048] On the one hand, the laboratory data blockchain generates a unique hash value for laboratory data, processing results, analysis reports, etc. through the hash algorithm and stores it in the laboratory data blockchain. Due to the immutable nature of the laboratory data blockchain, any modification to the laboratory data will cause a change in the hash value, ensuring that the data cannot be tampered with or forged throughout its life cycle. On the other hand, it generates an accurate timestamp for each data and records all operation processes to form a traceable audit record. All collection, processing, analysis, and verification operations of laboratory data will be accurately recorded on the chain, ensuring that the specific time and responsible person of each operation can be traced later. On the third hand, the smart contract in the laboratory data blockchain can automatically execute operations according to predefined rules, and can automatically trigger alarms or start the automatic adjustment and maintenance processes of equipment;

[0049] The IPFS distributed storage system is used to store laboratory data.

[0050] Embodiment 2, a data processing method for laboratory quality management based on artificial intelligence proposed by the present invention, which is applicable to the data processing system for laboratory quality management based on artificial intelligence proposed in Embodiment 1, specifically includes the following implementation steps:

[0051] S1. The laboratory management system (LIMS) collects personnel data, various types of equipment data, and sensor data (such as temperature, humidity, and pressure sensors) in the laboratory. For data that cannot be automatically obtained (such as manually recorded experimental conditions or results), the user manually inputs the data through the laboratory management system interface;

[0052] It should be noted that the data types include but are not limited to equipment parameters (temperature, pressure, operating status), environmental data, experimental result data, personnel data, etc. Each data contains detailed metadata (timestamp, equipment ID, operator ID, etc.);

[0053] S2. The laboratory management system aligns data from different sources through timestamps to ensure that all data can be analyzed and processed within the same time frame. It stores laboratory data at multiple nodes and generates a unique hash value (with a length of 256 bits) for each piece of collected laboratory data using the SHA-256 algorithm. The hash value of the data and related metadata (such as timestamps, data source devices, operators, etc.) are stored on the laboratory data blockchain. The laboratory data blockchain automatically adds an accurate timestamp to each blockchain operation to ensure that the data collection time can be accurately traced in the future. The specific content includes:

[0054] S2-1. The laboratory management system uses timestamps to synchronize data from different sources and merges data from different sources into a unified data structure to ensure that all data fields are correctly aligned;

[0055] S2-2. The laboratory management system stores laboratory data in a decentralized data storage system. This invention uses the IPFS (InterPlanetary File System) distributed storage system. After the data is stored in the IPFS distributed storage system, the system generates a unique IPFS hash;

[0056] S2-3. The laboratory management system calculates the hash value of the laboratory data and packages the data hash value and metadata (timestamp, IPFS hash) into a block through the smart contract Register(), as Figure 2 shown. Each block contains the following content:

[0057] Block header: Generate a block header, including version number, parent block, timestamp, nonce, and target hash;

[0058] Block body: Package the generated data hash value and IPFS hash into the block body;

[0059] Merkle root: Construct a Merkle tree, form a tree structure with data hash values (hash), and calculate the root hash value;

[0060] S2-4. Submit the block through the laboratory data blockchain network. The nodes of the laboratory data blockchain network use the ProofofWork consensus mechanism to verify the validity of the block. Once the block is verified, add the block to the end of the laboratory data blockchain to become part of the laboratory data blockchain. All laboratory data blockchain nodes update their copies of the laboratory data blockchain to ensure data synchronization;

[0061] S2-5. Record the timestamps of data and blocks in UTC time format to ensure the consistency of all time records. Generate accurate timestamps through the system clock or Network Time Protocol (NTP), and store them in the laboratory data blockchain;

[0062] S2-6. The database within the laboratory stores the hash values of data, and its structure includes: the hash values of test data, other metadata (including but not limited to timestamps, data types);

[0063] It should be noted that when a user needs to access laboratory data, the user sends an access request to the laboratory data blockchain network based on the hash value of the test data in the laboratory internal database. The nodes on the laboratory data blockchain query the records on the laboratory data blockchain according to the request, obtain the IPFS hash of the laboratory data and the corresponding laboratory data hash value, and use the IPFS hash retrieved from the laboratory data blockchain to access the actual laboratory data in the IPFS distributed storage system;

[0064] The user independently chooses whether to recalculate the hash value of the retrieved data, compare it with the hash value stored on the laboratory data blockchain, and then confirm whether the laboratory data has been tampered with;

[0065] S3. The laboratory management system uses the Z-score normalization method to process laboratory data and transmits the laboratory data to the AI analysis platform. The AI model in the AI analysis platform performs the following feature extraction, outputs the analysis results, and inputs the analysis results into the smart contract within the laboratory data blockchain;

[0066] Time series feature extraction: As Figure 3 shown, construct a time series feature extraction model, and extract time series features such as trends, seasonal variations, and periodic fluctuations from the device operation data through an LSTM (Long Short-Term Memory) network;

[0067] In the figure, (1) Input layer: Use the time series data as the input sequence of the model. Suppose there are N groups of input information in total, and the dimension of each group of information is D, then the input sequence is denoted as X = {x 1 , x 2 ,..., x N} ∈ R D×N ;

[0068] (2) Attention layer: Perform weight allocation according to the attention scores of the input sequence and each element of the output. First, calculate the attention distribution w = (w 1 , w 2 ,..., w N) Then, combine the attention distribution and the input vector to obtain a new vector as the input of the LSTM layer:

[0069] The calculation formula for attention is as follows:

[0070]

[0071] In the formula, w i (i = 1, 2,..., N) is the attention distribution, α is the attention variable, and α ∈ [1, N], i is the vector index, X is the input sequence, and s(x n , q) is the attention scoring function: s(x n , q) = (x n ) T q, and n = 1, 2,..., N;

[0072] (3) LSTM hidden layer: Use the LSTM neural network to mine the information in the time series of the output of the Attention layer, establish the mapping relationship between the multi-dimensional input sequence and the prediction target, and use the historical sequence as the input to predict the output;

[0073] (4) Fully connected layer: Perform dimensional transformation on the results of the LSTM hidden layer to obtain a one-dimensional vector;

[0074] (5) Output layer: Output the prediction results after dimensionality reduction by the fully connected layer;

[0075] Frequency domain feature extraction: Use the Fourier transform to convert the laboratory data from the time domain to the frequency domain, and extract potential periodic features, such as detecting the frequency pattern in the vibration data of the device;

[0076] Image feature extraction: For the image data generated in the experiment (such as microscope images, material defect detection images), use a convolutional neural network (CNN) to extract key features, such as edges, shapes, textures, etc. The present invention uses VGG16 as the network structure of the CNN;

[0077] Text feature extraction: As Figure 4 shown, construct a text feature extraction model, and through natural language processing (NLP) technology, use the BERT (preprocessing model) + BiLSTM (bidirectional long short-term memory network) + self attention (self-attention) + CRF (conditional random field) model to perform keyword extraction or topic model analysis on unstructured text data (such as experiment reports), and automatically identify important information;

[0078] It should be noted that before the laboratory management system transmits laboratory data to the AI analysis platform, the laboratory management system and the AI analysis platform need to conduct preliminary interactions. The laboratory management system checks the qualification of the AI analysis platform to analyze data. The interaction process is as follows:

[0079] (1) The AI analysis platform sends a data analysis request Query to the laboratory management system;

[0080] (2) When the laboratory management system receives the Query request, it generates a random number Rt and sends it to the AI analysis platform;

[0081] (3) The AI analysis platform generates a random number Rr and calculates the verification information I = H(ID||Rr||Rt). Then, it transmits the random number Rr and the verification information I to the laboratory management system;

[0082] where ID is the number of the AI analysis platform, || is the string concatenation symbol, and H is the hash function;

[0083] (4) The laboratory management system performs operations on the authorized AI analysis platform numbers ID in the laboratory internal database in a traversal manner to determine whether there is an ID i (1≦i≦n, where n is the number of authorized AI analysis platforms in the laboratory internal database) such that the equation H(ID i ||Rr||Rt) = I holds. If it exists, it means that the AI analysis platform has been authorized to analyze laboratory data. Then, the laboratory management system outputs the laboratory data to the AI model in the AI analysis platform;

[0084] S4. As shown in Table 1, set up a smart contract on the laboratory data blockchain to define the quality control rules for laboratory data, monitor the laboratory data in real time. When receiving the analysis results from the AI analysis platform, the smart contract automatically executes corresponding operations according to the rules and automatically approves the test data that needs to be approved and verified according to the predefined approval rules. Then, upload the approval results of each time to the laboratory data blockchain. The specific content includes:

[0085] S4-1. Conduct data verification in the smart contract Monitor(), check whether the data meets the standards defined in the contract, and the smart contract Execute() automatically executes the predefined operations according to the verification results. When abnormal data is detected, the system automatically takes measures according to the smart contract rules and records all trigger conditions and notifications on the laboratory data blockchain for subsequent traceability and auditing;

[0086] Table 1 Main description of the smart contract

[0087]

[0088] For example, the analyzed temperature trend is transmitted to the smart contract function. When the temperature exceeds the set threshold, the laboratory management system is notified through a laboratory data blockchain event (Event), and the laboratory management system sends an alarm notification to the device terminals of laboratory personnel (such as mobile phones and computers).

[0089] S4-2. The smart contract Examine() defines the approval process rules. When the set conditions are triggered (such as equipment failure), the system automatically generates an approval task and assigns the approval task to relevant personnel (such as laboratory administrators and quality control experts). The approval task includes, but is not limited to, checking abnormal situations and confirming data accuracy.

[0090] S4-3. Notify the assigned personnel of the generated approval task and require them to perform the approval. The approval request is generated through the smart contract and includes the data of the trigger condition and the recommended operation. The approver views the approval request and makes a decision (such as confirming the failure and approving the repair). The approval result is recorded and chained through the smart contract.

[0091] S4-4. Record the approval operation including the approver, approval time, and approval result on the laboratory data blockchain to generate an approval record, ensuring the transparency and compliance of the approval process.

[0092] Example 3. A data processing method for laboratory quality management based on artificial intelligence proposed by the present invention. Compared with Example 2, it further includes a method for sharing laboratory data, and the specific implementation steps are as follows:

[0093] S1. The laboratory manager classifies the collected laboratory data through the laboratory management system and establishes an access level information table for the laboratory, as shown in Table 2:

[0094] Table 2 Laboratory access level information

[0095]

[0096] S2. Attach attributes to the above-classified laboratory data to construct an attribute set Ω:

[0097]

[0098] Among them, the attribute Att 1 represents public data, the attribute Att 2 represents restricted data, and the attribute Att 3 represents sensitive data;

[0099] S3. Construct the association relationship between the attribute and the hash value of the laboratory data As shown in Table 3, establish a laboratory data information table stored on the laboratory data blockchain;

[0100] Table 3 Laboratory data information stored on the laboratory data blockchain

[0101]

[0102] S4. When a user or institution needs to submit an access request, stating the type of data required for access, the smart contract Access() on the laboratory data blockchain approves the data access request according to the predefined authorization rules, checks the identity, role permissions, and data attributes of the accessor, and confirms whether the type of data to be accessed meets the authorization conditions:

[0103] (1) When the attribute of the laboratory data applied to be viewed by the accessor is Att 1 At this time, the smart contract Access() directly approves, and the nodes on the laboratory data blockchain query the records on the laboratory data blockchain according to the request, obtain the IPFS hash of the laboratory data corresponding to the laboratory data hash value, and use the IPFS hash retrieved from the laboratory data blockchain to access the actual laboratory data in the IPFS distributed storage system;

[0104] (2) When the attribute of the laboratory data applied to be viewed by the accessor is Att 2 At this time, if the access level of the accessor is II, the smart contract Access() approves, allowing the accessor to view the data. The nodes on the laboratory data blockchain query the records on the laboratory data blockchain according to the request, obtain the IPFS hash of the laboratory data corresponding to the laboratory data hash value, and use the IPFS hash retrieved from the laboratory data blockchain to access the actual laboratory data in the IPFS distributed storage system;

[0105] Conversely, the accessor does not have the permission to access the data it wants to access, and the smart contract Access() does not approve and rejects its access behavior;

[0106] (3) When the attribute of the laboratory data applied to be viewed by the accessor is Att 3 At this time, if the access level of the accessor is III, the smart contract Access() approves, allowing the accessor to view the data. The nodes on the laboratory data blockchain view the records on the laboratory data blockchain according to the request, obtain the IPFS hash of the laboratory data applied to be accessed by the accessor, and use the IPFS hash retrieved from the laboratory data blockchain to access the laboratory data that the accessor wants to access in the IPFS distributed storage system;

[0107] On the contrary, if the accessing party does not have the permission to access the data it intends to access, the smart contract Access() approval fails, and the access behavior of the accessing party is rejected.

[0108] The embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited thereto, and various changes can be made without departing from the spirit of the present invention within the scope of knowledge possessed by those skilled in the art.

Claims

1. A data processing method for laboratory quality management based on artificial intelligence, characterized in that: The implementation steps include: S1. The laboratory management system collects laboratory data. For data that cannot be automatically obtained, the user manually enters the data through the laboratory management system interface; S2. The laboratory management system aligns data from different sources through timestamps. The laboratory data is stored in a decentralized data storage system. The SHA-256 algorithm is used to generate a unique hash value for each piece of collected laboratory data. The hash value of the data and related metadata are stored on the laboratory data blockchain. The laboratory data blockchain automatically adds an accurate timestamp for each on-chain operation. S3. The laboratory management system uses the Z-score standardization method to process laboratory data and transmits the laboratory data to the AI ​​analysis platform that has passed the data analysis qualification check. The AI ​​model in the AI ​​analysis platform extracts features, outputs analysis results, and inputs the analysis results into the smart contract in the laboratory data blockchain; The data analysis qualification verification process is as follows: A1. The AI ​​analysis platform sends a data analysis request Query to the laboratory management system; A2. When the laboratory management system receives the data analysis request Query, it generates a random number Rt and sends it to the AI ​​analysis platform; A3. The AI ​​analysis platform generates a random number Rr and calculates the verification information I=H(ID||Rr||Rt), and then transmits the random number Rr and the verification information I to the laboratory management system; Among them, ID is the ID of the AI ​​analysis platform, || is the string connection symbol, and H is the hash function; A4. The laboratory management system performs operations on the IDs of the authorized AI analysis platforms in the laboratory internal database in a traversal manner to determine whether a certain ID exists. i , so that H(ID i ||Rr||Rt)=I is true. If it exists, it means that the AI ​​analysis platform has been authorized to analyze the laboratory data. After that, the laboratory management system outputs the laboratory data to the AI ​​model in the AI ​​analysis platform; Where i is the serial number of the ID of the AI ​​analysis platform, 1≦i≦n, and n is the number of authorized AI analysis platforms in the laboratory's internal database; S4. Set up smart contracts on the laboratory data blockchain, define the quality control rules for laboratory data, and monitor laboratory data in real time. After receiving the analysis results from the AI ​​analysis platform, the smart contract automatically performs corresponding operations according to the rules, and automatically approves the test data that needs to be approved and verified according to the predefined approval rules, and then uploads each approval result to the laboratory data blockchain.

2. The data processing method for laboratory quality management based on artificial intelligence according to claim 1, characterized in that: The decentralized data storage system uses the IPFS distributed storage system. After the data is stored in the IPFS distributed storage system, the system generates a unique IPFS hash.

3. The data processing method for laboratory quality management based on artificial intelligence according to claim 1 is characterized in that: Laboratory data includes one or more of the following: personnel data, various equipment data, sensor data, and manually recorded experimental conditions or experimental results in the laboratory.

4. The data processing method for laboratory quality management based on artificial intelligence according to claim 1, characterized in that: The specific implementation steps of S4 are as follows: S4-1. Perform data verification in the smart contract to check whether the data meets the standards defined in the contract. The smart contract automatically executes the predetermined operation based on the verification results. When abnormal data is detected, the system automatically takes measures according to the smart contract rules and records all trigger conditions and notifications on the laboratory data blockchain. S4-2. Smart contracts define approval process rules. When the set conditions are triggered, approval tasks are automatically generated and assigned to management personnel. The approval tasks include checking for abnormalities and confirming data accuracy. S4-3. Notify the assigned personnel of the generated approval task and ask them to approve it. The approval request is generated through a smart contract, including the data of the triggering conditions and the recommended operations. The approver reviews the approval request and makes a decision. The approval result is recorded and uploaded to the chain through the smart contract. S4-4. Record the approval operations including the approval personnel, approval time, and approval results on the laboratory data blockchain to generate approval records to ensure the transparency and compliance of the approval process.

5. The data processing method for laboratory quality management based on artificial intelligence according to claim 1 is characterized in that: The AI ​​model in the AI ​​analysis platform performs feature extraction. The extracted features include: using the constructed time series feature extraction model to extract time series features from equipment operation data through the LSTM network, using Fourier transform to convert laboratory data from the time domain to the frequency domain to extract potential periodic features, using the constructed convolutional neural network to extract image features from image data generated in the experiment, and using the constructed text feature extraction model to perform keyword extraction and topic model analysis on unstructured text data through natural language processing technology using the BERT+BiLSTM+self attention+CRF model.

6. The data processing method for laboratory quality management based on artificial intelligence according to claim 5 is characterized in that: The lab data can be accessed as follows: S7-1. The laboratory manager classifies the collected laboratory data through the laboratory management system and establishes the relationship between data types and permission levels: <public data, permission level I>, <restricted data, permission level II> and <sensitive data, permission level III>; S7-2. Attach attributes to the classified laboratory data to construct the attribute set Ω: ； Among them, attribute Att1 represents public data, attribute Att2 represents restricted data, and attribute Att3 represents sensitive data; S7-3. Construct the association between attributes and laboratory data hash values ​​{attributeAtt k , laboratory data hash value k }; S7-4. When a user or organization needs to submit an access request, indicating the type of data required for access, the smart contract on the laboratory data blockchain will review and approve the data access request based on the pre-defined authorization rules, check the identity, role authority, and data attributes of the accessing party, and then k , laboratory data hash value k }, confirm whether the accessed data type meets the authorization conditions; S7-5, if the data access request of the accessing party is approved, the node on the laboratory data blockchain queries the record on the laboratory data blockchain according to the request, obtains the IPFS hash of the laboratory data corresponding to the laboratory data hash value, and uses the IPFS hash found on the laboratory data blockchain to access the actual laboratory data in the IPFS distributed storage system.

7. A data processing system for laboratory quality management based on artificial intelligence, which is applicable to a data processing method for laboratory quality management based on artificial intelligence according to any one of claims 1 to 6, characterized in that: include: Laboratory management system for collecting and preprocessing laboratory data; AI analysis platform, the running carrier is an AI algorithm box, which has a series of built-in AI models to analyze various types of laboratory data; The laboratory data blockchain generates a unique hash value for the laboratory data through a hash algorithm, stores it in the laboratory data blockchain, generates an accurate timestamp for each data, and records all operation processes to form a traceable audit record. The smart contracts in the laboratory data blockchain can automatically perform operations according to predefined rules, and can automatically trigger alarms or start automatic adjustment and maintenance processes for equipment; IPFS distributed storage system, used to store laboratory data.

Images