AI auxiliary sample sharing method and system

Abstract

The invention discloses an AI auxiliary sample sharing method and system. The AI auxiliary sample sharing method comprises the following steps: searching in a dialogue form between a GUI (Graphical User Interface) and a system; the AI intelligent matching module is used for carrying out semantic analysis on a query request submitted by a user based on a deep learning algorithm, and intelligently screening clinical special disease biological samples meeting user requirements from a clinical special disease biological sample database; personalized recommendation of the user can be provided according to historical behaviors and preferences of the user and currently inquired context information, and matching precision and efficiency are improved; by adopting the method, the efficiency of inquiring the required clinical special disease biological sample by clinical personnel can be greatly improved, so that the overall clinical research test efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and particularly relates to an AI-assisted sample sharing method and system. The platform aims to integrate extensive sample data resources and utilize advanced AI technology to achieve efficient and intelligent clinical disease-specific biological sample data retrieval function, thereby assisting users in quickly searching and obtaining the required clinical disease-specific biological sample data in the form of dialogue to support AI model construction and optimization in various application scenarios. BACKGROUND

[0002] With the rapid development of big data and artificial intelligence technology, clinical disease-specific biological sample data plays a crucial role in AI model construction. However, traditional clinical disease-specific biological sample data retrieval methods often rely on keyword search or complex query statements, which may be too cumbersome and inefficient for non-professional users. In addition, existing clinical disease-specific biological sample sharing platforms mostly lack intelligent retrieval mechanisms and are difficult to provide accurate clinical disease-specific biological sample recommendations according to users' actual needs. SUMMARY

[0003] To solve the above technical problems, the present application proposes an AI-assisted sample sharing method and system. The system achieves its purpose through the following technical features, including:

[0004] Platform architecture, including a clinical disease-specific biological sample database, an AI retrieval engine, a user interaction interface, and a background management system. The clinical disease-specific biological sample database stores a large amount of clinical disease-specific biological sample data, covering various data types and formats. The AI retrieval engine is the core of the platform, utilizing machine learning algorithms and deep learning models to index and classify sample data. The user interaction interface provides an intuitive and easy-to-use operation interface, supporting sample search in the form of dialogue. The background management system is used to maintain the normal operation of the platform, including data updating, user permission management, etc.

[0005] AI retrieval method, users input natural language queries or dialogue requests through the user interaction interface. The AI retrieval engine receives user requests, performs semantic understanding and intent recognition, extracts the clinical disease-specific biological sample features that the user wants to query, searches for matching clinical disease-specific biological sample data in the clinical disease-specific biological sample database according to the extracted features, generates a candidate clinical disease-specific biological sample list, sorts the candidate clinical disease-specific biological samples using a sorting algorithm, places the clinical disease-specific biological samples that best meet user needs at the front, and the user interaction interface displays the sorted sample list and provides detailed information viewing.

[0006] The dialogue interaction function supports natural language dialogue in the user interaction interface, and the user can interact with the platform in the form of dialogue. The platform can intelligently recommend and provide personalized services according to the user's feedback and context information. The user can request the platform to provide relevant clinical special disease biological sample data interpretation, usage suggestions or other support information through dialogue.

[0007] Security and privacy protection, the platform uses data encryption technology to ensure the security of clinical special disease biological sample data during transmission and storage, implements strict access control mechanism, only authorized users can access specific clinical special disease biological sample data, provides data desensitization and anonymization processing function, to protect user privacy and data security.

[0008] Technical effects, compared with the traditional clinical special disease biological sample data retrieval method, the present application has the following advantages: it improves the efficiency and accuracy of clinical special disease biological sample data retrieval, reduces the operation difficulty of the user; through the intelligent retrieval mechanism, the function of accurate recommendation according to the user demand is realized; an intuitive and easy-to-use dialogue interaction interface is provided, which enhances the user experience and satisfaction; the safety and privacy protection measures are strengthened, and the safety and reliability of the clinical special disease biological sample data are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to ensure the purpose, technical solution and advantages of the embodiments of the present application to be clearly exhibited, the technical solutions in the embodiments of the present application will be described in detail and systematically below in combination with the drawings of the embodiments. It should be clear that the described embodiments only represent a part of the examples of the present application, rather than the whole scope. These examples are intended to explain the present application, rather than to limit its application scope. Based on the disclosed embodiments of the present application, all other embodiments that can be deduced by those skilled in the art without creative labor are considered to fall within the protection scope of the present application.

[0010] For those skilled in the art, unless otherwise defined, all terms (including technical terms and scientific terms) used in this text are consistent with the meanings commonly understood by those skilled in the art to which the embodiments of the present application belong. At the same time, it should be understood that terms such as those defined in general dictionaries should be given the same meaning as in the context of the prior art, and unless specifically defined in this text, these terms should not be interpreted in an idealized or overly formal manner.

[0011] Figure 1 is a flowchart of an AI-assisted sample sharing method and system in the embodiments of the present application; DETAILED DESCRIPTION

[0012] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of the present application.

[0013] With reference to Figure 1 The embodiments of the present application provide a technical solution: an AI-assisted sample sharing method and system, comprising the following steps: collecting clinical special disease biological sample data, the collected data coming from multiple channels; preprocessing the clinical special disease biological sample data; storing the clinical special disease biological sample data; user interaction for data query and feedback through dialogue natural language; user input natural language processing (NLP) to extract important features of user input data; training a BERT model to extract data features stored in the library; extracting and matching the data to be searched by the user; returning the clinical special disease biological sample data to be searched by the user; user input personalized recommendation; storage data security and privacy.

[0014] The shared clinical special disease biological sample data comes from multiple channels, including scientific research institutions, medical institutions, enterprises, etc. Due to the inconsistent storage methods and data formats of different channel data sources, the data will be cleaned, de-duplicated, standardized before data query to ensure the quality and consistency of the data. In addition, the database uses a relational database or a non-relational database to store the clinical special disease biological sample data before storing the data in the database. According to the data characteristics and query requirements, a suitable storage structure is selected, and an index and cache mechanism is designed to improve the data retrieval efficiency.

[0015] The user input and natural language processing (NLP) are described. The user interface inputs natural language queries or dialogue requests through a graphical user interface (GUI). The user interface focuses on user experience, and provides friendly input prompts and immediate feedback based on the examples provided by the user.

[0016] Specifically, after the AI retrieval engine receives the user input, it first performs word segmentation, dividing the user input instance into a combination of several characters or fragments to help better understand the meaning of the user input instance. In addition, part-of-speech tagging is performed during word segmentation to determine whether each word is a noun, verb, adjective, or other part of speech. At the same time, data preprocessing is performed to convert the original text into a form that can be better understood and processed by a computer, removing noise and redundant information in the text to improve the quality and readability of the text.

[0017] The training BERT model, preparing a large amount of training sample data, the clinical special disease biological sample data contains the following characteristics: sample type, sample name, sample capacity, patient information, disease information, topic, etc. Load the pre-trained BERT model using the Transformers library of Hugging Face, input the training data in batches, and finally achieve semantic understanding and intent recognition of the deep learning model (BERT). Extract the clinical special disease biological sample features that the user wants to query, including sample type, sample name, sample capacity, patient information, disease information, topic, and other key information.

[0018] After extracting the clinical special disease biological sample features of the user input instance, the feature vector is stored in the inverted index structure, the data in the library is quickly retrieved, the similarity between the user query and the feature vector in the sample database is evaluated, the Jaccard similarity is used, the ratio between the obtained user clinical special disease biological sample feature set and the library clinical special disease biological sample data feature set is calculated, and the Top-K method is used: return the K clinical special disease biological samples with the highest similarity.

[0019] After obtaining the above K clinical special disease biological samples with the highest similarity, the machine learning model is used to arrange the clinical special disease biological sample that best meets the user's needs in the front and display it on the user's graphical user interface (GUI).

[0020] Specifically, the above-mentioned machine learning model sorts the candidate clinical special disease biological samples, and a large amount of clinical special disease biological sample feature data is prepared in advance for training. The trained clinical special disease biological sample feature data contains the following characteristics: sample type, sample name, sample capacity, patient information, disease information, topic, etc. The model learns these data and gradually masters how to generate sorting weights based on features.

[0021] After the user inputs the instance, the user's query history and preferences are used to provide personalized recommendations using collaborative filtering. Improve user satisfaction and platform efficiency, and introduce natural language generation (NLG) technology to realize intelligent question and answer and interaction with the user, answer the user's questions about clinical special disease biological samples, and provide query suggestions, etc.

[0022] The system data data security and privacy protection, a secure communication channel is established between the server and the client, all the transmitted data are transmitted by using the SSL / TLS protocol for sample data encryption, ensure that the data is not stolen or tampered with in the transmission process, and use a secure key management scheme to ensure the security of the key generation, storage, distribution and update process. Data storage generates a unique encryption key for each piece of data and stores it in a secure key management system. The access and use of the key are strictly controlled. And the clinical special disease biological sample data stored on the server is encrypted, using symmetric encryption or asymmetric encryption algorithm, according to the importance and sensitivity of the data, select the appropriate encryption strength. In addition, the backup and recovery process of encrypted data also needs to be kept in an encrypted state, to ensure that the data is not leaked in the backup and recovery process.

[0023] Specifically, the user access platform port needs to pass through user verification, using multi-factor authentication methods such as username + password, mobile phone verification code, biometric identification, etc., to ensure the authenticity of the user's identity. Only authenticated users can access the platform. And the user access system permission is determined according to the role assigned by the system, the system assigns different permissions to each different role, such as data access permission, data editing permission, data deletion permission, etc. Only users with corresponding permissions can access or operate specific clinical special disease biological sample data, in addition, the system records the user's access and operation log, including access time, access content, operation type, etc. Information, and use machine learning algorithms to analyze log data, detect abnormal access and operation behavior, and timely discover and handle potential security risks.

[0024] Specifically, when storing data, sensitive data is desensitized, such as replacing the real name with a pseudonym, replacing the ID number with a partially hidden or fake number, etc., to prevent important data from being leaked, and clear privacy protection policies, clear privacy protection measures for data collection, use, storage and sharing, etc.

[0025] The following is based on the preferred embodiment of the present application, through the foregoing detailed description, those skilled in the related art can fully implement various changes and adjustments without departing from the basic technical concept of the present application. The technical protection scope of the present application is not limited to the specific description in the specification, but should be defined according to the scope defined in the claims.

Claims

1. An AI-assisted sample sharing method and system, characterized in that, Includes the following steps: The user interaction provides a user-friendly GUI interface, enabling users to submit clinical disease-specific biological sample query requests, browse clinical disease-specific biological sample details, initiate clinical disease-specific biological sample sharing requests, and manage personal information through natural language dialogue; the module also supports real-time feedback, which can instantly display the results of AI intelligent matching and the status of clinical disease-specific biological sample sharing. Sample management is responsible for the collection, editing, review, and storage of clinical disease-specific biological sample information, ensuring the accuracy, completeness, and timeliness of clinical disease-specific biological sample data; this module also has sample classification and label management functions, making it convenient for users to quickly retrieve information based on key information such as sample type, source, and disease type; AI-powered intelligent matching is based on deep learning algorithms. It performs semantic parsing on user-submitted queries and intelligently filters out clinical disease-specific biological samples that meet the user's needs from a clinical disease-specific biological sample database. This module can also continuously optimize the matching algorithm based on the user's historical behavior, preferences, and the context of the current query, thereby improving matching accuracy and efficiency. The access control system assigns different access and operation permissions to users based on their identity, role, and needs. This module also supports dynamic permission adjustment, which can adjust the user's permission scope in real time based on changes in user behavior, project needs, or organizational strategies. The data storage and processing module is responsible for storing key data such as user information, clinical disease-specific biological sample information, query records, and shared records, and provides efficient data processing and analysis capabilities. This module also has data backup and recovery functions to ensure data security and reliability.

2. The AI-assisted sample sharing method and system according to claim 1, characterized in that, The user interaction module also includes intelligent question answering, which can understand user questions and provide accurate answers based on natural language processing technology; the intelligent question answering also supports multi-turn dialogue function, which can perform contextual association based on user feedback and provide more in-depth services.

3. The AI-assisted sample sharing method and system according to claim 1, characterized in that, The AI ​​intelligent matching module also has a user intent recognition function, which can accurately determine the user's query intent, such as finding specific types of clinical disease biological samples, comparing the differences between different clinical disease biological samples, etc. Based on user intent, the module will provide more accurate matching results and recommendations.

4. The AI-assisted sample sharing method and system according to claim 1, characterized in that, The permission management module also has fine-grained permission control, which can assign different access and operation permissions to users for specific clinical disease biological samples, datasets or functional modules; this helps to protect the security of sensitive data and prevent unauthorized access and operation.

5. The AI-assisted sample sharing method and system according to claim 1, characterized in that, The data storage and processing module also has data encryption and desensitization functions, which can encrypt and store sensitive data and desensitize it when needed; this helps protect user privacy and data security and ensure the compliant use of data.

6. The AI-assisted sample sharing method and system according to claim 1, characterized in that, The platform also features intelligent recommendation capabilities, which can recommend relevant clinical disease-specific biological samples, datasets, or expert resources based on users' historical behavior, preferences, and current needs. This helps users quickly find the resources they need, improving research efficiency and quality.

Images