AI optimal storage system having automatic storage function and capable of being connected with hospital HIS system

Abstract

The invention discloses an AI optimal storage system with an automatic storage function and capable of being connected with a hospital HIS system, and relates to the technical field of medical information management and data storage, the AI optimal storage system comprises a data input end, a data processing end, an HIS system docking module, an automatic storage strategy generation end, a cache management end, a data security and authority control module and a user interface end; the data input end is responsible for receiving and collecting medical data from all systems of a hospital, and the medical data comprise operation images, medical images, patient electronic medical records and monitoring data. The data processing end is used for preprocessing input data, and the preprocessing comprises format conversion, data cleaning, compression and duplicate removal. The intelligent degree of data storage is remarkably improved, remarkable effects are achieved in the aspects of data management efficiency, cross-system integration, postoperative risk management, data safety and the like, and the high-standard requirements of modern hospitals for operation data storage, analysis and management are met.

Description

Technical Field

[0001] The present invention relates to the field of medical information management and data storage technology, and in particular to an AI optimal storage system with an automatic storage function that can be connected to a hospital HIS system. Background Art

[0002] In the modern medical environment, efficient storage and management of medical data has become the key to the development of hospital informatization. With the popularization of medical equipment and the digitization of medical processes, hospitals generate massive amounts of data every day, including surgical videos, medical images, patient medical records, etc. These data are not only large in volume but also require real-time, traceability and high security. However, traditional hospital information systems (HIS) are mainly used to manage basic medical information such as medical records, drugs, and finances. They lack intelligent support for the management and storage of a large number of multimedia files (such as videos and images), especially in terms of automatic data storage, dynamic storage optimization, and cross-departmental data sharing.

[0003] In addition, traditional medical data storage methods often use fixed storage strategies and cannot perform intelligent storage grading based on data usage frequency or importance. This method easily wastes storage space and is difficult to meet the needs of real-time data retrieval.

[0004] Therefore, there is an urgent need for an AI storage system that can automatically optimize data storage strategies and support seamless integration with the hospital HIS system to change this situation. Summary of the invention

[0005] The purpose of this invention is to solve the shortcomings of the prior art and propose an AI storage system with automatic storage function that can be connected to the hospital HIS system. Its advantages are that it not only significantly improves the intelligence of data storage, but also achieves remarkable results in data management efficiency, cross-system integration, postoperative risk management and data security, meeting the high standards of modern hospitals for surgical data storage, analysis and management.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The AI ​​storage system with automatic storage function can be connected to the hospital HIS system, including data input terminal, data processing terminal, HIS system docking module, automatic storage strategy generation terminal, cache management terminal, data security and permission control module and user interface terminal;

[0008] The data input end is responsible for receiving and collecting medical data from various hospital systems, including surgical images, medical images, patient electronic medical records and monitoring data; the data processing end is used to pre-process the input data, including format conversion, data cleaning, compression and deduplication, to provide standardized and simplified data for subsequent storage optimization; the input image and video data are processed in a unified format and stored in common video and image formats; the deduplication algorithm is used to detect duplicate data and remove redundant content; and large file data is losslessly compressed to reduce storage requirements;

[0009] The HIS system docking module is seamlessly connected with the hospital information system, so that surgical images, medical images and patient medical records can be synchronized to the HIS system in real time; it provides standardized data interfaces and APIs, provides services in the form of RESTful interfaces, and synchronizes data with the HIS system;

[0010] The automatic storage strategy generation end is used to automatically generate storage strategies based on data usage frequency, importance and key frames, and perform hierarchical management of hot and cold data and dynamic hierarchical storage; based on deep learning algorithms, data features are analyzed to classify data into key frames and noise frames to achieve hierarchical storage and resource optimization; the cache management end is used to optimize the system's response speed and data access efficiency, and perform multi-level cache management on high-frequency access data;

[0011] The data security and authority control module is responsible for authority control and security protection of medical data in the system; the user interface provides a data visualization management platform to display data status, storage strategy and access log for monitoring and management.

[0012] The present invention is further configured such that the data input terminal supports multiple data sources and data formats, is compatible with operating room equipment, imaging equipment and patient information systems, and automatically detects and identifies different types of data; uses the DICOM protocol and the HL7 protocol to ensure compatibility with medical imaging and hospital management systems and data standardization; performs integrity checks when receiving data, and uses verification or hash algorithms to ensure accurate data transmission.

[0013] The present invention is further configured that the data cleaning adopts a machine learning algorithm, and the machine learning algorithm formula is: The machine learning algorithm automatically identifies abnormal data and removes it during the cleaning process.

[0014] The present invention is further configured such that the real-time synchronization of information adopts a data synchronization and update mechanism, and the data synchronization and update mechanism realizes real-time data synchronization with the HIS system through timed polling or event triggering, so that the latest surgical images and medical record data can be viewed in the HIS system; it supports two-way communication with the HIS system, and data is interconnected and shared between the optimal storage system and the HIS system, and different data types are converted and parsed, so that medical data can be efficiently accessed in each system.

[0015] The present invention is further configured such that the automatic storage strategy generation module dynamically adjusts the storage location of the data in combination with the frequency of data use, access time and content importance, and stores high-frequency data on a fast-access storage medium (SSD) and low-frequency data on a lower-cost medium (HDD); the system sets storage priorities for different data, and important surgical data and imaging data are preferentially stored in fast-speed storage media to ensure rapid reading of critical data.

[0016] The present invention is further configured such that the cache management end adopts a three-level cache mechanism, namely, hot cache, cold cache and remote cache, to ensure that high-frequency data is always in the cache; uses Redis high-efficiency memory database to implement hot cache and quickly respond to user requests; performs distributed cache management between memory, SSD and main storage, respectively, and stores frequently accessed hot data on faster storage media.

[0017] The present invention is further configured such that the data security and permission control module is based on multi-factor authentication (MFA), including user name and password, fingerprint verification, and dynamic password for user authentication, and uses the RBAC (role-based access control) model to set permissions for different roles; and uses AES and RSA encryption algorithms to encrypt sensitive data in storage and transmission.

[0018] The present invention is further configured such that the data security and permission control module records access logs of all users, including access time and operation content, and supports auditing and tracing of access history; combines AI to analyze user behavior patterns, dynamically adjusts permissions based on risk assessment, and prohibits abnormal or malicious operation behaviors.

[0019] The present invention is further configured such that the user interface adopts a modern UI framework to design a data management interface, and charts are used to display the storage status, usage frequency and storage medium occupancy of the data; a monitoring dashboard is provided to display the system's storage status, data cache hit rate, access frequency and response speed key performance indicators in real time; and data search, classification, deletion and permission setting are supported.

[0020] The beneficial effects of the present invention are:

[0021] 1. The AI ​​storage system with automatic storage function can be connected to the hospital HIS system. The system uses AI-driven automatic storage strategies to perform intelligent hierarchical storage according to data usage frequency, importance, size and other attributes, making high-frequency data easier to retrieve quickly, while low-frequency data is reasonably stored to save space; the AI ​​storage system can be seamlessly connected to the hospital HIS system to achieve real-time sharing of surgical data, imaging materials, reports and other information, making it easier for doctors to call data across departments and systems, thereby improving diagnosis and treatment efficiency;

[0022] 2. The AI ​​storage system with automatic storage function can be connected to the hospital HIS system. The system can automatically record and trace key data and operations throughout the operation, including the time of each step, the use of operating tools, intraoperative physiological monitoring data, etc., to facilitate postoperative analysis and clinical research; the system strictly controls data access rights, records the time and operator of each data access, ensures the security of surgical data during transmission, storage, and call, and meets compliance requirements; the AI ​​storage system has good scalability, supports future technology and data growth needs, and supports remote maintenance and intelligent monitoring, simplifying the system maintenance process;

[0023] 3. This AI storage system with automatic storage function can be connected to the hospital HIS system. The system not only significantly improves the intelligence level of data storage, but also has achieved remarkable results in data management efficiency, cross-system integration, postoperative risk management and data security, meeting the high standards of modern hospitals for surgical data storage, analysis and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a schematic diagram of the overall structure of the AI ​​storage system with automatic storage function that can be connected to the hospital HIS system proposed by the present invention;

[0025] Figure 2 This is a schematic diagram of the data processing module of the AI ​​storage system with automatic storage function and connectable to the hospital HIS system proposed by the present invention;

[0026] Figure 3 This is a block diagram of the automatic storage strategy generation of the AI ​​optimal storage system with automatic storage function that can be connected to the hospital HIS system proposed by the present invention. DETAILED DESCRIPTION

[0027] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0028] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0029] Reference Figure 1-3 , an AI storage system with automatic storage function that can be connected to the hospital HIS system, including data input terminal, data processing terminal, HIS system docking module, automatic storage strategy generation terminal, cache management terminal, data security and permission control module and user interface terminal;

[0030] The data input terminal is responsible for receiving and collecting medical data from various hospital systems, including surgical images, medical images, patient electronic medical records and monitoring data; ensuring that the system can fully obtain the required input information; the data input terminal supports multiple data sources and data formats, is compatible with operating room equipment, imaging equipment and patient information systems, and automatically detects and identifies different types of data; uses DICOM protocol and HL7 protocol to ensure compatibility with medical imaging and hospital management systems and data standardization; performs integrity verification when receiving data, and uses verification or hash algorithms to ensure accurate data transmission;

[0031] The data processing end is used to pre-process the input data, including format conversion, data cleaning, compression and deduplication, to provide standardized and streamlined data for subsequent storage optimization; the input image and video data are processed in a unified format and stored in common video and image formats; deduplication algorithms are used to detect duplicate data and remove redundant content; large file data is losslessly compressed to reduce storage requirements; efficient compression formats such as H.265 are used to maintain image clarity while reducing space usage; data cleaning uses machine learning algorithms, and the machine learning algorithm formula is: Machine learning algorithms automatically identify abnormal data and remove it during the cleaning process;

[0032] The HIS system docking module is seamlessly connected with the hospital information system, so that surgical images, medical images and patient medical records can be synchronized to the HIS system in real time; ensuring the consistency of data in different systems; providing standardized data interfaces and APIs, providing services in the form of RESTful interfaces, and synchronizing with HIS system data; the interface design follows the HL7 standard to ensure the standardized transmission of medical data;

[0033] Real-time information synchronization adopts a data synchronization and update mechanism. The data synchronization and update mechanism realizes real-time data synchronization with the HIS system through timed polling or event triggering. The latest surgical images and medical record data can be viewed in the HIS system; it supports two-way communication with the HIS system. Data is shared between the Youcun system and the HIS system, and different data types are converted and analyzed, so that medical data can be efficiently accessed in each system.

[0034] The HIS system docking module ensures real-time data synchronization with the hospital's HIS system and supports two-way data interaction to ensure data consistency.

[0035] Integration of RESTful API and message queue: Design interfaces through RESTful API to achieve real-time data transmission, and use message queues (Apache Kafka) to ensure asynchronous two-way transmission of data and achieve real-time data updates.

[0036] The incremental update mechanism based on database triggers automatically triggers data synchronization every time data is updated in the HIS system, and the changes are synchronized to this system to avoid data consistency issues.

[0037] API integration and incremental updates: Cross-system data access is achieved through API interfaces, and incremental data synchronization is achieved in combination with database triggers, reducing the pressure brought by data updates.

[0038] The automatic storage strategy generation end is used to automatically generate storage strategies based on data usage frequency, importance and key frames, and perform hierarchical management of hot and cold data and dynamic hierarchical storage; based on deep learning algorithms, data features are analyzed and data is classified into key frames and noise frames to achieve hierarchical storage and resource optimization; the automatic storage strategy generation module dynamically adjusts the storage location of the data based on data usage frequency, access time and content importance, storing high-frequency data on fast-access storage media (SSD) and low-frequency data on lower-cost media (HDD); the system sets storage priorities for different data, and important surgical data and imaging data are preferentially stored in fast storage media to ensure fast reading of key data.

[0039] An unsupervised learning method is adopted to summarize the video with a deep summarization network (DSN). The whole process is a sequential decision-making process. DSN is an encoding-decoding structure, in which the encoder is a convolutional neural network that can extract video frame features, and the decoder is a bidirectional LSTM network that can generate probabilities based on the selected actions.

[0040] The encoder of DSN is a convolutional neural network that extracts visual features t=1T from the input video frame t=1T of length T. The decoder is a bidirectional recurrent neural network (BiRNN) with a fully connected layer on top. After the extracted t=1T is input to the decoder, the corresponding hidden state t=1T is generated. Each ht is the concatenation of the previous hidden state htf and the next hidden state htb. In practice, GoogLeNet is used as the CNN model, and LSTM training is used to improve the performance of RNN.

[0041] During training, DSN receives a reward R(S) to evaluate the generated summary. The goal of DSN is to continuously generate high-quality video summaries to maximize the reward. Usually, high-quality video summaries must be both representative and colorful. To achieve this goal, the researchers proposed a novel reward method, which consists of a diversity reward Rdiv and a representative reward Rrep.

[0042] With this reward, the agent can select the frame closest to the cluster center of the feature space. Finally, Rdiv and Rrep work together to guide DSN learning: R(S) = Rdiv + Rrep

[0043] The system storage space is divided into cold data area and hot data area. Based on AI prediction results, high-frequency access data (such as key frames of surgical videos) are stored in the hot data area (SSD), and low-frequency data is stored in the cold data area (HDD).

[0044] Through AI models, storage data is regularly re-evaluated and data with low access frequency is migrated from hot areas to cold areas to save hot storage space.

[0045] The cache management end is used to optimize the system's response speed and data access efficiency, and perform multi-level cache management on frequently accessed data. The cache management end adopts a three-level cache mechanism, namely hot cache, cold cache and remote cache, to ensure that high-frequency data is always in the cache. The Redis efficient memory database is used to implement hot cache and quickly respond to user requests. Distributed cache management is performed between memory, SSD and main storage, and frequently accessed hot data is stored on faster storage media.

[0046] Cache replacement algorithm: Use cache replacement algorithms such as LRU (least recently used) and LFU (least frequently used) to achieve dynamic cache updates and automatic data elimination to ensure efficient cache utilization. Cache data update: Set a refresh mechanism for cache data. When the data in the main storage is updated, the data in the cache is also updated in real time to ensure data consistency. Cache hit rate optimization: Combine AI algorithms to predict user access behavior to data, load data with high likelihood of access in advance, and improve cache hit rate.

[0047] The data security and permission control module is responsible for the permission control and security protection of medical data within the system; the data security and permission control module is based on multi-factor authentication (MFA), including username and password, fingerprint verification, and dynamic password for user authentication, and uses the RBAC (role-based access control) model to set permissions for different roles; AES and RSA encryption algorithms are used to encrypt sensitive data in storage and transmission; the data security and permission control module records the access logs of all users, including access time and operation content, and supports auditing and tracing of access history; combined with AI to analyze user behavior patterns, permissions are dynamically adjusted based on risk assessment, and abnormal or malicious operations are prohibited.

[0048] The user interface provides a visual management platform for data, displaying data status, storage strategy and access log for monitoring and management; the user interface adopts a modern UI framework to design the data management interface, with charts showing the data storage status, usage frequency and storage media occupancy; provides a monitoring dashboard to display the system's storage status, data cache hit rate, access frequency and response speed key performance indicators in real time; supports data search, classification, deletion and permission setting.

[0049] When storage resources are approaching their limits or abnormal access occurs, the system sends reminders or alarm notifications to administrators through the interface to ensure the security and stability of the system.

[0050] These modules together form an AI optimal storage system with intelligent storage, real-time data sharing and data security, meeting the hospital's multiple needs for medical data storage, management and access.

[0051] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The AI ​​storage system with automatic storage function can be connected to the hospital HIS system, which is characterized by: It includes data input terminal, data processing terminal, HIS system docking module, automatic storage strategy generation terminal, cache management terminal, data security and authority control module and user interface terminal; The data input end is responsible for receiving and collecting medical data from various hospital systems, including surgical images, medical images, patient electronic medical records and monitoring data; the data processing end is used to pre-process the input data, including format conversion, data cleaning, compression and deduplication, to provide standardized and simplified data for subsequent storage optimization; the input image and video data are processed in a unified format and stored in common video and image formats; the deduplication algorithm is used to detect duplicate data and remove redundant content; and large file data is losslessly compressed to reduce storage requirements; The HIS system docking module is seamlessly connected with the hospital information system, so that surgical images, medical images and patient medical records can be synchronized to the HIS system in real time; it provides standardized data interfaces and APIs, provides services in the form of RESTful interfaces, and synchronizes data with the HIS system; The automatic storage strategy generation end is used to automatically generate storage strategies based on data usage frequency, importance and key frames, and perform hierarchical management of hot and cold data and dynamic hierarchical storage; based on deep learning algorithms, data features are analyzed to classify data into key frames and noise frames to achieve hierarchical storage and resource optimization; the cache management end is used to optimize the system's response speed and data access efficiency, and perform multi-level cache management on high-frequency access data; The data security and authority control module is responsible for authority control and security protection of medical data in the system; the user interface provides a data visualization management platform to display data status, storage strategy and access log for monitoring and management.

2. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 1 is characterized in that: The data input terminal supports multiple data sources and data formats, is compatible with operating room equipment, imaging equipment and patient information systems, and automatically detects and identifies different types of data; uses DICOM protocol and HL7 protocol to ensure compatibility with medical imaging and hospital management systems and data standardization; performs integrity verification when receiving data, and uses verification or hash algorithms to ensure accurate data transmission.

3. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 1 is characterized in that: The data cleaning adopts a machine learning algorithm, and the formula of the machine learning algorithm is: The machine learning algorithm automatically identifies abnormal data and removes it during the cleaning process.

4. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 3 is characterized in that: The real-time synchronization of information adopts a data synchronization and update mechanism, which realizes real-time data synchronization with the HIS system through timed polling or event triggering. The latest surgical images and medical record data can be viewed in the HIS system; it supports two-way communication with the HIS system, and data is shared between the Youcun system and the HIS system, and different data types are converted and analyzed, so that medical data can be efficiently accessed in each system.

5. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 4 is characterized in that: The automatic storage strategy generation module dynamically adjusts the storage location of the data based on the frequency of data use, access time and content importance, storing high-frequency data on fast-access storage media (SSD) and low-frequency data on lower-cost media (HDD); the system sets storage priorities for different data, with important surgical data and imaging data preferentially stored in fast-speed storage media to ensure fast reading of critical data.

6. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 1 is characterized in that: The cache management end adopts a three-level cache mechanism, namely hot cache, cold cache and remote cache, to ensure that high-frequency data is always in the cache; Use Redis efficient memory database to implement hot cache and quickly respond to user requests; Distributed cache management is performed between memory, SSD and main storage respectively, and frequently accessed hot data is stored on faster storage media.

7. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 1 is characterized in that: The data security and permission control module is based on multi-factor authentication (MFA), including username and password, fingerprint verification, and dynamic password for user authentication, and uses the RBAC (role-based access control) model to set permissions for different roles; and uses AES and RSA encryption algorithms to encrypt sensitive data in storage and transmission.

8. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 7 is characterized in that: The data security and permission control module records the access logs of all users, including access time and operation content, and supports auditing and tracing of access history; combines AI to analyze user behavior patterns, dynamically adjusts permissions based on risk assessment, and prohibits abnormal or malicious operations.

9. The AI ​​storage system with automatic storage function and connectable to the hospital HIS system according to claim 1 is characterized in that: The user interface adopts a modern UI framework to design a data management interface, and charts display the data storage status, usage frequency and storage medium occupancy; provides a monitoring dashboard to display the system's storage status, data cache hit rate, access frequency and response speed key performance indicators in real time; Supports data search, classification, deletion and permission setting.

Images