Artificial intelligence model management system based on metadata

Abstract

The invention discloses an artificial intelligence model management system based on metadata, which relates to the technical field of model management, and comprises a metadata processing module for acquiring metadata including continuous metadata and discrete metadata, calculating a field center point code of the continuous metadata into a label distribution vector, counting the total number of categories of the discrete metadata, and calculating the total number of categories of the discrete metadata; coding is carried out to construct a mapping matrix, vector splicing is carried out, and linear mapping is carried out by using a linear transformation matrix to convert into a latent vector with a uniform length. According to the system, Gaussian kernel label distribution coding of continuous metadata and coding of discrete metadata are carried out, the complete range of variables can be covered through detailed segments, and it is ensured that different types of features can be fused in the same potential space through unified metadata latent vectors; and meanwhile, the comparability of the features is ensured, and the diffusion process of step-by-step denoising is introduced into the model, so that medical data generated by reasoning each time has higher diversity.

Description

Technical Field

[0001] The present invention relates to the technical field of model management, and in particular to an artificial intelligence model management system based on metadata. Background Art

[0002] Metadata, as the "data" that describes data, contains information such as data source, data type, data relationship, data processing method, etc. With the continuous advancement of big data technology, enterprises and scientific research institutions are increasingly relying on efficient metadata management systems when processing massive amounts of information. In the medical field, prediction and analysis based on metadata are also gradually being proposed and used. These systems can not only improve data query efficiency, but also reveal the deep-level value hidden behind the data through metadata correlation analysis. However, the existing metadata management and AI model training methods still have certain shortcomings. Most of the current mainstream technologies focus on the processing and modeling of a single type of metadata. The difference processing between continuous metadata and discrete metadata is often not sufficiently refined. In particular, it is difficult to efficiently capture the potential relationship between categories for medical information, resulting in information loss and limited model expression capabilities. In addition, existing methods usually ignore how to efficiently integrate heterogeneous data sources. Especially in the case of multimodal data in the medical field, it is difficult for model training to simultaneously consider the richness and completeness of the data. Therefore, existing technologies face large performance bottlenecks and application limitations when processing complex data types and large-scale data. Summary of the invention

[0003] In view of the above existing problems, the present invention is proposed.

[0004] Therefore, the present invention provides an artificial intelligence model management system based on metadata to solve the problem that the difference processing between continuous metadata and discrete metadata is often not sufficiently refined, especially for medical information, it is difficult to efficiently capture the potential relationship between categories, resulting in information loss and limited model expression ability. In addition, existing methods usually ignore how to efficiently integrate heterogeneous data sources, especially in the case of multimodal data in the medical field, and model training is difficult to simultaneously consider the richness and completeness of the data.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an artificial intelligence model management system based on metadata, comprising: The metadata processing module collects metadata including continuous metadata and discrete metadata, encodes the center point of the continuous metadata calculation field into a label distribution vector, counts the total number of discrete metadata categories, encodes and constructs a mapping matrix, performs vector concatenation, and uses a linear transformation matrix to perform linear mapping and convert into a uniform length latent vector; Medical data processing module, extracts shallow vectors of medical images, and fuses them with unified latent vectors to generate conditional vectors, introduces ControlNet control module according to external structural features, generates control vectors, fuses them with unified latent vectors of metadata to generate conditional vectors, gradually adds noise and builds diffusion model according to U-Net network, performs reverse denoising training, and generates restored latent vectors; Model deployment module: The client deploys the diffusion model to standardize data, aggregates model parameters to update the global model, and evaluates the global model; The query module analyzes the uncertainty of the global model, defines the knowledge graph to record the model operation data, uploads the operation log, converts the query statements entered by the user and performs data query.

[0006] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the vector concatenation is performed by using a linear transformation matrix to perform linear mapping into a uniform length latent vector, including: For each field of continuous metadata, the center point of each field is calculated based on the variable range and score segment of the field sample value; Construct a Gaussian kernel label distribution function to encode its value into a vector, and map the input value x of the field to a label distribution vector; For each discrete metadata field, count the number of unique values ​​of all samples in the field as the total number of categories. , perform Embedding encoding to construct a mapping matrix with a dimension of , d is a fixed embedding dimension; Each row represents the embedding vector of a category, and the embedding vector representation of each discrete variable is obtained; The concatenated label distribution vector and the embedded vector representation of discrete metadata have an output vector dimension of p for all continuous variables and an embedded vector dimension of d for all discrete variables. and the total number of discrete variables , define the dimension D after splicing; The encoding vectors of all fields are sequentially concatenated into a total vector, and a linear transformation matrix is ​​used to perform linear mapping to convert it into a latent vector of uniform length.

[0007] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the generated control vector is fused with the unified latent vector of the metadata to generate a conditional vector, including: According to the corresponding medical image data of the clinical data, the image data is paired with the unified latent vector of the metadata, and a pre-trained image encoder is used to extract the shallow vector of the medical image data and fuse it with the unified latent vector of the metadata to generate a conditional vector; Determine the external structural features of medical image data, including key points, labels, and structural information of the image, introduce the ControlNet control module, take the condition vector and external structural features as input, and generate the control vector , and fused with the image latent vector to generate a fused latent vector; The forward diffusion process is performed by gradually adding noise based on the image latent vector; A diffusion model is constructed based on the U-Net network, and reverse denoising training is performed to restore the latent vector of the original image from the noise state. The mean square error loss MSE is used to train the model to predict noise; Use the Adam optimizer to perform gradient descent optimization and update the model parameters. If the model loss no longer decreases significantly during the continuous iteration, the iteration is stopped. The trained diffusion model is able to generate a conditionally generated restored latent vector based on the noisy input and the conditional vector in the latent space.

[0008] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the aggregation model parameters update the global model, including: Collect customer metadata based on customer terminals, use the metadata calibrator Meta-Calibrator to standardize the metadata of each client, deploy the diffusion model and use local data for model training; A weighted coefficient is determined for each client, and the model parameters of all clients are aggregated using the weighted average method, and the global model is updated based on the global model parameters.

[0009] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the global model evaluation includes: Perform FID and SSIM evaluation on the global model; The sum of the mean and standard deviation of the historical distribution difference between the generated image and the real image is used as the difference threshold. If the distribution difference value is less than or equal to the difference threshold, it is judged that the original features of the image are retained. The sum of the mean and standard deviation of the historical structural difference values ​​between the generated image and the real image is used as the structural threshold. If the structural difference value is greater than or equal to the structural threshold, it is judged that the original features of the image are retained.

[0010] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the uncertainty of the global model analysis model includes: A repeated sampling mechanism LAS is introduced for the global model to perform multiple random initializations on the input image latent vector and conditional vector, and obtain the corresponding restored latent vector, and calculate the average prediction value and average standard deviation to generate an uncertainty score; The sum of the mean and twice the standard deviation of the historical uncertainty score is used as the scoring threshold. If the uncertainty score is less than the scoring threshold, the uncertainty is judged to be low.

[0011] As a preferred solution of the metadata-based artificial intelligence model management system described in the present invention, the collected metadata includes continuous metadata and discrete metadata, including: By traversing each column field in the original clinical data table, reading the value type and value distribution range of each column, if the field value is a continuous numeric type, it is marked as continuous metadata, if the field value is a discrete classification label, it is marked as discrete metadata, and each row in the clinical data table is used as a sample.

[0012] As a preferred solution of the metadata-based artificial intelligence model management system described in the present invention, the definition of the knowledge graph to record model operation data refers to using the DataHub architecture to define the triple structure of the knowledge graph KG, recording the model's input / output mapping, data source, processing steps, model parameters and execution environment information.

[0013] As a preferred solution of the metadata-based artificial intelligence model management system described in the present invention, the uploading operation log refers to recording the operation log for each operation step of the model and storing it as structured metadata.

[0014] As a preferred solution of the metadata-based artificial intelligence model management system of the present invention, wherein: the conversion of the query statement input by the user and the data query include: Use the BERT model to convert the query statements entered by the user into SPARQL statements to query the metadata in the knowledge graph; Based on the user query, the system will return the corresponding model path, inference conditions, and detailed information of input data.

[0015] The beneficial effects of the present invention are as follows: by performing Gaussian kernel label distribution encoding of continuous metadata and encoding of discrete metadata, the full range of variables can be covered through detailed segmentation, and different types of features can be integrated into the same latent space through a unified metadata latent vector, while ensuring the comparability of features. A gradual denoising diffusion process is introduced into the model, so that the medical data generated by each inference has higher diversity. The introduction of noise enables the model to generate a variety of latent vector representations, thereby enhancing the flexibility of the model in actual clinical applications, especially in disease progression prediction, personalized medicine and simulation of different clinical trials, providing more prediction options. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1 This is a schematic diagram of the structure of the metadata-based artificial intelligence model management system in Example 1; Figure 2 This is a flow chart of the metadata-based artificial intelligence model management system in Example 1. DETAILED DESCRIPTION

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0019] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0020] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0021] Example 1, reference Figure 1 to Figure 2 , which is the first embodiment of the present invention, and provides an artificial intelligence model management system based on metadata, comprising the following steps: S1, collect metadata including continuous metadata and discrete metadata, encode the center point of the continuous metadata calculation field into a label distribution vector, count the total number of categories of discrete metadata, encode and construct a mapping matrix, perform vector concatenation, and use a linear transformation matrix to perform linear mapping to convert into a uniform length latent vector; Preferably, the collected metadata includes continuous metadata and discrete metadata, including: By traversing each column field in the original clinical data table, reading the value type and value distribution range of each column, if the field value is a continuous numeric type, it is marked as continuous metadata, if the field value is a discrete classification label, it is marked as discrete metadata, and each row in the clinical data table is used as a sample.

[0022] By traversing the original clinical data table and classifying it according to the value type and value distribution range of each column, the data can be effectively and automatically marked as continuous or discrete metadata, which helps to clearly identify the data structure and provide accurate data type identification for subsequent data processing, modeling and analysis, ensuring consistency and efficiency in the processing process. At the same time, it helps the model to adopt appropriate processing methods according to the characteristics of the data when processing different types of data, thereby improving the performance and accuracy of the model.

[0023] Furthermore, vector concatenation is performed and linear transformation matrices are used to perform linear mapping into latent vectors of uniform length, including: For each field of continuous metadata, the center point of each field is calculated based on the variable range and score segment of the field sample value, expressed as: ; in represents the center point of the i-th Gaussian kernel, n represents the number of segments, b represents the maximum value of the variable range, and a represents the minimum value of the variable range; Construct a Gaussian kernel label distribution function to encode its value into a vector and map the input value x of the field to a label distribution vector, expressed as: ; in represents the label distribution vector of the i-th field, represents the smoothing factor, which can be calculated according to the formula Determine, used to ensure that the distribution coverage width is consistent; For each discrete metadata field, count the number of unique values ​​of all samples in the field as the total number of categories. , perform Embedding encoding to construct a mapping matrix with a dimension of , d is a fixed embedding dimension; Each row represents the embedding vector of a category, and the embedding vector representation of each discrete variable is obtained. For each sample, the category index j is obtained by looking up the table, and the jth row is extracted as the embedding vector; The concatenated label distribution vector and the embedded vector representation of discrete metadata have an output vector dimension of p for all continuous variables and an embedded vector dimension of d for all discrete variables. and the total number of discrete variables , define the concatenated dimension D, expressed as: ; The encoding vectors of all fields are sequentially concatenated into a total vector, and a linear transformation matrix is ​​used to perform linear mapping to convert it into a latent vector of uniform length. This vector is a metadata conditional representation and will be passed into the AI ​​model control path, expressed as: ; ; in Represents the original vector after all metadata are concatenated, with a length of D. express, A unified latent vector representation representing metadata, represents the linear transformation matrix, which can be initialized by the standard normal distribution and then fixed during training.

[0024] By performing Gaussian kernel label distribution encoding for continuous metadata and Embedding encoding for discrete metadata, the full range of variables can be covered through detailed segmentation, avoiding information loss caused by single numerical input, so that the model can more accurately learn the underlying laws of the data when processing these features. For discrete data, Embedding encoding can effectively reduce the dimension while retaining the category information by mapping each category to a dense vector space, thereby improving the generalization ability and prediction accuracy of the model; The unified metadata latent vector ensures that different types of features can be integrated into the same latent space, while ensuring the comparability of the features. The relative relationship between the features can be captured by the model, thus providing a more comprehensive conditional input for the model. This latent vector representation not only improves the integration of the data, but also enhances the model's comprehensive understanding of various input features, thereby further improving the effect of generation tasks or prediction tasks.

[0025] S2, extract the shallow vector of the medical image and fuse it with the unified latent vector to generate a conditional vector, introduce the ControlNet control module according to the external structural features, generate a control vector, fuse it with the unified latent vector of the metadata to generate a conditional vector, gradually add noise and build a diffusion model according to the U-Net network, perform reverse denoising training, and generate a restored latent vector; Preferably, a control vector is generated and fused with a unified latent vector of metadata to generate a conditional vector, including: According to the corresponding medical image data of the clinical data, the image data is paired with the unified latent vector of the metadata, and the shallow vector of the medical image data is extracted using the pre-trained image encoder and fused with the unified latent vector of the metadata to generate a conditional vector, which is expressed as: ; in represents the conditional vector, represents the image latent vector output by the image encoder, represents the fusion coefficient, which is determined based on historical experience; Determine the external structural features of medical image data, including key points, labels, and structural information of the image, introduce the ControlNet control module, take the condition vector and external structural features as input, and generate the control vector , and fused with the image latent vector to generate a fused latent vector, expressed as: ; in represents the fused latent vector, represents the control weight, which is obtained by training and optimizing based on the calibrated training data; Based on the image latent vector, the forward diffusion process is performed by gradually adding noise, making each image latent vector more blurred, which is expressed as: ; in and denote the fused latent vectors at time steps t and t-1 respectively, represents the decay factor at step t (determined based on historical experience), represents standard Gaussian noise, with covariance being the noise of the identity matrix; A diffusion model is constructed based on the U-Net network, and reverse denoising training is performed to restore the latent vector of the original image from the noise state. The mean square error loss MSE is used to train the model to predict noise, which is expressed as: ; in represents the calculation loss, Represents the U-Net network model used to predict noise; Use the Adam optimizer to perform gradient descent optimization and update the model parameters. If the model loss no longer decreases significantly during the continuous iteration, the iteration is stopped. The trained diffusion model is able to generate a conditionally generated restored latent vector based on the noise input and conditional vector in the latent space, and reconstruct it into image data through the decoder.

[0026] By pairing the unified latent vector of medical data and metadata, the generated conditional vector can provide more comprehensive input information for the model, so that the model can consider the latent characteristics of medical data and the individualized metadata of patients at the same time. By controlling the conditional vector through the ControlNet module, the model's fine-grained control over medical data can be further enhanced. Especially in the task of generating medical data, controlling the influence of external structural features can help the model generate medical data that is more in line with medical knowledge, such as accurately reconstructing the anatomical structure of a specific part, or adjusting certain features of medical data (such as lesion areas). Through optimized control weights, the model can more accurately adjust the generation process based on the feedback of training data to ensure that the generated images are more in line with clinical needs. By performing a forward diffusion process on the fused latent vector and adding noise, the model learns how to gradually recover from a fuzzy state, so that it can better generate high-quality medical data when facing unknown or fuzzy inputs. The introduction of the diffusion model can also further improve the diversity of generated medical data and allow the model to generate different versions of medical data, enhancing the model's exploration ability. The diffusion model built on the U-Net network can gradually recover clear image latent vectors from noise through reverse denoising training. The model can accurately predict image details during the denoising process while maintaining an efficient training process. The use of the Adam optimizer ensures that the model parameters can be updated stably, thereby accelerating convergence and improving model performance. The trained diffusion model can generate a restored latent vector based on the noise input and the conditional vector, and convert it into image data through the decoder. By combining medical image data and metadata to generate a conditional vector, the patient's individual information can be effectively introduced into the image generation process. The introduction of the control module (ControlNet) allows the image generation process to take into account external structural features and can accurately control specific anatomical structures during the generation process, effectively reducing the distortion of the generated image and improving the medical credibility of the image. The introduction of a gradual denoising diffusion process in the model makes the medical data generated in each inference more diverse. The introduction of noise enables the model to generate diverse latent vector representations, thereby outputting different versions of medical data, enhancing the flexibility of the model in actual clinical applications, especially in disease progression prediction, personalized medicine and simulation of different clinical trials, providing more prediction options.

[0027] S3, the client deploys the diffusion model to standardize the data, aggregates the model parameters to update the global model, and evaluates the global model; Preferably, aggregating model parameters to update the global model includes, Collect customer metadata based on customer terminals, use the metadata calibrator Meta-Calibrator to standardize the metadata of each client, deploy the diffusion model and use local data for model training; Determine the weight coefficient for each client, and use the weighted average method to aggregate the model parameters of all clients, and update the global model according to the global model parameters, which is expressed as: ; in represents the global model parameters, Indicates the total number of clients. represents the data volume of client i, D represents the total data volume, represents the parameters of the i-th client model, including the weights and biases of the diffusion model, As model weights for different clients.

[0028] By using a metadata calibrator, we can ensure that the data of all clients are processed under the same standard, thereby improving the uniformity and consistency of the data. Deploying a diffusion model and using local data for training can enable each client to make personalized adjustments based on its own local data, thereby improving the model performance of each client. Using the weighted average method to aggregate the model parameters of each client helps to better balance the influence of each client.

[0029] Further, the global model evaluation includes, Perform FID and SSIM evaluation on the global model; The sum of the mean and standard deviation of the historical distribution difference between the generated image and the real image is used as the difference threshold. If the distribution difference value is less than or equal to the difference threshold, it is judged that the original features of the image are retained. The sum of the mean and standard deviation of the historical structural difference values ​​between the generated image and the real image is used as the structural threshold. If the structural difference value is greater than or equal to the structural threshold, it is judged that the original features of the image are retained.

[0030] By using FID and SSIM to evaluate the difference between generated images and real images, the improvement and fidelity of image quality can be effectively measured. The FID evaluation can quantify the distribution difference between generated images and real images, while the SSIM evaluation focuses on the structural similarity of images.

[0031] S4, analyzes the uncertainty of the global model, defines a knowledge graph to record model operation data, uploads operation logs, converts the query statements entered by users and performs data query; Preferably, the global model is analyzed for model uncertainty, including, A repeated sampling mechanism LAS is introduced for the global model to perform multiple random initializations on the input image latent vector and conditional vector. The output of the model under different initialization conditions is obtained through multiple samplings, thereby obtaining the uncertainty information of the result, and obtaining the corresponding restored latent vector. The average prediction value and average standard deviation are calculated to generate an uncertainty score. The sum of the mean and twice the standard deviation of the historical uncertainty score is used as the scoring threshold. If the uncertainty score is less than the scoring threshold, the uncertainty is judged to be low.

[0032] By introducing the repeated sampling mechanism (LAS), the model can obtain outputs under different initialization conditions by randomly initializing the input image latent vector and conditional vector multiple times, thereby better reflecting the uncertainty of the generated results; By calculating the average prediction value and standard deviation of the restored latent vector obtained through multiple sampling, an uncertainty score can be generated for each output, which can effectively identify which generated results have higher stability and which have greater uncertainty.

[0033] Furthermore, the knowledge graph is defined to record model operation data, including: The DataHub architecture is used to define the triple structure of the knowledge graph KG, recording the model's input / output mapping, data source, processing steps, model parameters, and execution environment information. Each triple contains: entities (such as datasets, models, and inference results), attributes (such as data types, calculation methods, and model status), and relationships (such as the mapping relationship between models and datasets).

[0034] By using the DataHub architecture to define the triple structure of the knowledge graph (KG), the model's input / output mapping, data sources, processing steps, model parameters, and execution environment information can be systematically recorded and managed. This structured information storage method provides a clear framework for data traceability, management, and analysis, ensuring the transparency of data flow and model operation. The triple form makes the definition of each entity, attribute, and relationship clearer, facilitating collaboration and information sharing across systems or teams, thereby improving the traceability, maintainability, and efficiency of the workflow.

[0035] Furthermore, uploading operation logs means recording operation logs for each operation step of the model and storing them as structured metadata for auditing and review.

[0036] All metadata are indexed and associated with specific training processes, inference results, or abnormal events. Using the graph query interface, users can quickly find all operation histories related to a model or dataset. By using the DataHub architecture to define the triple structure of the knowledge graph (KG), the model's input / output mapping, data sources, processing steps, model parameters, and execution environment information can be systematically recorded and managed to ensure the transparency of data flow and model operation. The triple form makes the definition of each entity, attribute, and relationship clearer, facilitating collaboration and information sharing across systems or teams, thereby improving the traceability, maintainability, and efficiency of the workflow.

[0037] Further, the query statement input by the user is converted and data query is performed, including: Use the BERT model to convert the query statements entered by the user into SPARQL statements to query the metadata in the knowledge graph; Based on the user query, the system will return the corresponding model path, inference conditions and detailed information of input data; By using the BERT large model to convert user query statements into SPARQL statements, more natural and intelligent user interaction can be achieved, allowing users to obtain metadata-related information in the knowledge graph through simple queries. This automated conversion improves the flexibility and efficiency of queries, allowing the system to quickly return relevant model paths, reasoning conditions, and detailed information on input data based on user needs, thereby optimizing the application of the knowledge graph and user experience.

[0038] In summary, the present invention can cover the full range of variables through detailed segmentation by performing Gaussian kernel label distribution encoding of continuous metadata and encoding of discrete metadata, and ensure that different types of features can be integrated in the same latent space through a unified metadata latent vector, while ensuring the comparability of features. A gradual denoising diffusion process is introduced into the model, so that the medical data generated by each inference has higher diversity. The introduction of noise enables the model to generate diverse latent vector representations, thereby enhancing the flexibility of the model in actual clinical applications, especially in disease progression prediction, personalized medicine, and simulation of different clinical trials, providing more prediction options.

[0039] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An artificial intelligence model management system based on metadata, characterized in that: include: The metadata processing module collects metadata including continuous metadata and discrete metadata, encodes the center point of the continuous metadata calculation field into a label distribution vector, counts the total number of discrete metadata categories, encodes and constructs a mapping matrix, performs vector concatenation, and uses a linear transformation matrix to perform linear mapping and convert into a uniform length latent vector; Medical data processing module, extracts shallow vectors of medical images, and fuses them with unified latent vectors to generate conditional vectors, introduces ControlNet control module according to external structural features, generates control vectors, fuses them with unified latent vectors of metadata to generate conditional vectors, gradually adds noise and builds diffusion model according to U-Net network, performs reverse denoising training, and generates restored latent vectors; Model deployment module: The client deploys the diffusion model to standardize data, aggregates model parameters to update the global model, and evaluates the global model; The query module analyzes the uncertainty of the global model, defines the knowledge graph to record the model operation data, uploads the operation log, converts the query statements entered by the user and performs data query.

2. The metadata-based artificial intelligence model management system according to claim 1, characterized in that: The vector concatenation is performed by using a linear transformation matrix to perform linear mapping into a uniform length latent vector, including: For each field of continuous metadata, the center point of each field is calculated based on the variable range and score segment of the field sample value; Construct a Gaussian kernel label distribution function to encode its value into a vector, and map the input value x of the field to a label distribution vector; For each discrete metadata field, count the number of unique values ​​of all samples in the field as the total number of categories. , perform Embedding encoding to construct a mapping matrix with a dimension of , d is a fixed embedding dimension; Each row represents the embedding vector of a category, and the embedding vector representation of each discrete variable is obtained; The concatenated label distribution vector and the embedded vector representation of discrete metadata have an output vector dimension of p for all continuous variables and an embedded vector dimension of d for all discrete variables. and the total number of discrete variables , define the dimension D after splicing; The encoding vectors of all fields are sequentially concatenated into a total vector, and a linear transformation matrix is ​​used to perform linear mapping to convert it into a latent vector of uniform length.

3. The metadata-based artificial intelligence model management system according to claim 2, characterized in that: The generated control vector is fused with the unified latent vector of the metadata to generate a conditional vector, including: According to the corresponding medical image data of the clinical data, the image data is paired with the unified latent vector of the metadata, and a pre-trained image encoder is used to extract the shallow vector of the medical image data and fuse it with the unified latent vector of the metadata to generate a conditional vector; Determine the external structural features of medical image data, including key points, labels, and structural information of the image, introduce the ControlNet control module, take the condition vector and external structural features as input, and generate the control vector , and fused with the image latent vector to generate a fused latent vector; The forward diffusion process is performed by gradually adding noise based on the image latent vector; A diffusion model is constructed based on the U-Net network, and reverse denoising training is performed to restore the latent vector of the original image from the noise state. The mean square error loss MSE is used to train the model to predict noise. Use the Adam optimizer to perform gradient descent optimization and update the model parameters. If the model loss no longer decreases significantly during the continuous iteration, the iteration is stopped. The trained diffusion model is able to generate a conditionally generated restored latent vector based on the noisy input and the conditional vector in the latent space.

4. The metadata-based artificial intelligence model management system according to claim 3, characterized in that: The aggregation model parameters update the global model, including: Collect customer metadata based on customer terminals, use the metadata calibrator Meta-Calibrator to standardize the metadata of each client, deploy the diffusion model and use local data for model training; A weighted coefficient is determined for each client, and the model parameters of all clients are aggregated using the weighted average method, and the global model is updated based on the global model parameters.

5. The metadata-based artificial intelligence model management system according to claim 4, characterized in that: The global model evaluation includes: Perform FID and SSIM evaluation on the global model; The sum of the mean and standard deviation of the historical distribution difference between the generated image and the real image is used as the difference threshold. If the distribution difference value is less than or equal to the difference threshold, it is judged that the original features of the image are retained. The sum of the mean and standard deviation of the historical structural difference values ​​between the generated image and the real image is used as the structural threshold. If the structural difference value is greater than or equal to the structural threshold, it is judged that the original features of the image are retained.

6. The metadata-based artificial intelligence model management system according to claim 5, characterized in that: The global model analysis model uncertainty includes: A repeated sampling mechanism LAS is introduced for the global model to perform multiple random initializations on the input image latent vector and conditional vector, and obtain the corresponding restored latent vector, and calculate the average prediction value and average standard deviation to generate an uncertainty score; The sum of the mean and twice the standard deviation of the historical uncertainty score is used as the scoring threshold. If the uncertainty score is less than the scoring threshold, the uncertainty is judged to be low.

7. The metadata-based artificial intelligence model management system according to claim 6, characterized in that: The collected metadata includes continuous metadata and discrete metadata, including: By traversing each column field in the original clinical data table, reading the value type and value distribution range of each column, if the field value is a continuous numeric type, it is marked as continuous metadata, if the field value is a discrete classification label, it is marked as discrete metadata, and each row in the clinical data table is used as a sample.

8. The metadata-based artificial intelligence model management system according to claim 7, characterized in that: The definition of the knowledge graph to record model operation data refers to using the DataHub architecture to define the triple structure of the knowledge graph KG, recording the model's input / output mapping, data source, processing steps, model parameters and execution environment information.

9. The metadata-based artificial intelligence model management system according to claim 8, characterized in that: The uploading operation log refers to recording the operation log for each operation step of the model and storing it as structured metadata.

10. The metadata-based artificial intelligence model management system according to claim 9, characterized in that: The step of converting the query statement input by the user and performing data query includes: Use the BERT model to convert the query statements entered by the user into SPARQL statements to query the metadata in the knowledge graph; Based on the user query, the system will return the corresponding model path, inference conditions, and detailed information of input data.

Images