Medical equipment disinfection traceability management system

Abstract

The present invention discloses a medical device disinfection traceability management system, which relates to the field of data processing technology. The system includes: an initialization part, a disinfection process recording part, and a traceability and risk assessment part; the initialization part is used to generate a unique 256-bit identifier containing medical device information for each medical device, and use an improved Merkle-Patricia tree structure to build a medical device information library; the disinfection process recording part is used to use a zero-knowledge proof protocol to encrypt and record the disinfection process data to ensure the consistency and non-tamperability of the data written into the blockchain; the traceability and risk assessment part is used to build a graph neural network of the medical device information library, find the node most likely to have disinfection anomalies, and complete the traceability of medical devices with disinfection anomalies. The present invention aims to achieve safe and non-tamperable storage and traceability of disinfection records.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a medical device disinfection traceability management system. Background Art

[0002] With the rapid development of modern medical technology, medical devices play an increasingly important role in medical institutions. The use of medical devices covers all aspects of diagnosis, treatment, monitoring and rehabilitation, and their quality and safety are directly related to the health and treatment effect of patients. In order to ensure the safe use of medical devices, medical institutions must strictly abide by the standard operating procedures for disinfection and sterilization to ensure that each medical device is thoroughly disinfected before each use. However, the disinfection process of medical devices is complicated and involves many factors, such as the type of disinfectant, disinfection time, environmental conditions, etc. Negligence in any link may lead to incomplete disinfection, thereby increasing the risk of hospital infection. Therefore, the traceability and management of disinfection records have become an important link in ensuring the safe use of medical devices.

[0003] In the prior art, research on medical device disinfection management mainly focuses on automated disinfection equipment and recording systems for the disinfection process. For example, some medical institutions have begun to use intelligent disinfection equipment that can automatically detect disinfectant concentration, time, and temperature, and record disinfection process data in the local system. However, these disinfection records are usually managed as individual devices, lacking a unified, traceable system to integrate and verify the disinfection process of all devices. This limitation not only increases the complexity of management, but also lays hidden dangers for potential disinfection anomalies and equipment failures. Especially in modern hospital environments with huge amounts of disinfection data and a wide variety of equipment, the cost and complexity of manually checking and managing disinfection records have increased dramatically, making it difficult to effectively respond to actual needs.

[0004] In addition, the existing medical device disinfection management system also has major defects in the security and privacy protection of disinfection data. Medical device disinfection records often involve sensitive data such as operator information, disinfectant batches, equipment types and environmental conditions. Once these data are leaked or tampered with, they may pose a serious threat to the operation of the hospital and the privacy of patients. Although some systems have adopted basic encryption measures to protect data security, there is still a risk of information leakage or attack during data storage and transmission. In addition, in the existing systems, most disinfection records are stored in local servers or private clouds, lacking the ability to share data across institutions and the guarantee of non-tamperability. This architecture limits the traceability of medical device disinfection records. Once equipment disinfection anomalies or safety hazards occur, it is difficult to conduct comprehensive and reliable traceability and verification in a short period of time. Summary of the invention

[0005] The main purpose of the present invention is to provide a medical device disinfection traceability management system, which ensures the transparency and immutability of data through blockchain, optimizes data retrieval efficiency through Merkle-Patricia tree, protects data privacy through zero-knowledge proof protocol, and improves storage and transmission efficiency by combining compression algorithm. The present invention not only improves the security and privacy protection capabilities of disinfection records, but also significantly improves the storage efficiency of the system and the intelligent detection capabilities of abnormal disinfection, and is suitable for the disinfection management and traceability of large-scale medical equipment.

[0006] In order to solve the above problems, the technical solution of the present invention is achieved as follows:

[0007] A medical device disinfection traceability management system, the system comprising: an initialization part, a disinfection process recording part and a traceability and risk assessment part; the initialization part is used to generate a unique 256-bit identifier containing medical device information for each medical device, regard each medical device as a node, and use an improved Merkle-Patricia tree structure to construct a medical device information base based on the identifier corresponding to each node. In the medical device information base, each medical device is a node; the disinfection process recording part is used to use a zero-knowledge proof protocol to encrypt and record the disinfection process data to obtain a disinfection record, use a block sorting compression algorithm to compress the disinfection record, and then compress the compressed The compressed results are stored in a distributed file system, and the hash value, timestamp and corresponding medical device identifier of the compressed results in the distributed file system are written into the blockchain to ensure the consistency and non-tamperability of the data written into the blockchain; the traceability and risk assessment part is used to build a graph neural network for the medical device information library. In the graph neural network, the graph attention network layer is used to capture the relationships between medical devices, and abnormal node identification based on relationship anomalies is performed to find the node most likely to have disinfection anomalies. Then, through the zero-knowledge proof protocol, the hash value, timestamp and corresponding medical device identifier corresponding to the node are obtained from the blockchain to complete the traceability of medical devices with disinfection anomalies.

[0008] Furthermore, in the initialization part, the medical device information includes at least: device number , the length is bit; date of production , length is 8 bits; production batch , the length is bit; manufacturer code , the length is bit; device type code , the length is All medical device information is spliced ​​in a fixed order to obtain the initial information string ;in, Represents a string concatenation operation; uses a cryptographic hash function to concatenate the initial information string Perform hash calculation to generate a 256-bit hash value As an identifier: ; The hash algorithm has an output length of 256 bits; to ensure that the generated identifier Unique in the entire medical device information database, using the following uniqueness verification mechanism: define the hash set in the medical device information database as ,in is the number of medical devices; Indicates Identifier of a medical device; check the generated Do the following relations satisfy: ; If not satisfied, introduce a random salt value , regenerate a new hash value ; Repeat the above process until a unique .

[0009] Furthermore, when the initialization part uses the improved Merkle-Patricia tree structure to build the medical device information library, the hash value of the current medical device is compared. If there is a common prefix with the hash prefix of the existing nodes in the tree, determine the path position where the node should be inserted based on the longest common prefix; if there is no common prefix, insert a new node and adjust the tree structure so that it still maintains the compression characteristics of the Patricia tree; define each node The storage structure is a triple: ;in, Represents the common prefix from the root to the node; Indicates the hash value of the medical device corresponding to this node ; Represents the list of child nodes of the node; in order to maintain data integrity and verification, the hash value of each node in the improved Merkle-Patricia tree structure is generated by combining the hash values ​​of its child nodes, that is, the hash value of each internal node By the hash value of its child nodes Calculated by the following formula: , For the The hash value of the child nodes, is the total number of child nodes; in order to prevent the improved Merkle-Patricia tree structure from becoming too unbalanced after multiple insertions, affecting query efficiency, a dynamic balancing strategy is introduced to ensure the maximum depth of the improved Merkle-Patricia tree structure by introducing hash value space partitioning and path depth optimization algorithms Controlled in logarithmic time complexity In the range of The following optimization conditions are met: ;in, is the number of nodes in the current tree, is the pre-balancing factor, ; The current depth of the modified Merkle-Patricia tree structure.

[0010] Furthermore, in the disinfection process record section, the disinfection process data At least include: operator ID, equipment number of medical device, disinfection type, disinfectant name, disinfectant batch, disinfection equipment ID, equipment type, disinfection start time, disinfection end time, disinfection duration, ambient temperature and ambient humidity; map the disinfection process data to an irreversible hash value through a hash function It is a hash algorithm with an output length of 512 bits. The zero-knowledge proof protocol uses Pedersen commitment for encrypted recording. , the generated commitment value for: ; Among them, the commitment value It is a model The value calculated in the group is used for subsequent zero-knowledge proof verification. This commitment value ensures the privacy of the disinfection process data while allowing verification of the integrity and authenticity of the data; is a public parameter, Generators in a group; is a nonlinear term; is a random salt of the commitment value, used in Pedersen commitment to increase randomness to prevent direct deduction through brute force , each time a commitment value is generated, are all different, so that even the same You will also get different commitment values; is a preset prime number used for modular operations; As a disinfection record.

[0011] Furthermore, in the disinfection process record part, when the disinfection record is compressed using the block sorting compression algorithm, the disinfection record Split into blocks equal to the number of nodes in the modified Merkle-Patricia tree structure ; For each block Apply the block sort compression algorithm: ;in, is to make The position with the smallest lexicographic order among all cyclic shifts; Indicates the block compression algorithm; The resulting application moves to the front-end coding: ;in is the character currently being processed, It’s the Iverson bracket; For the characters processed; Indicates moving to the front-end for processing; The subscript index of the current character; further compression using improved arithmetic coding : ;in, It is The frequency of the symbol; is the arithmetic coding algorithm; the compressed result is: .

[0012] Furthermore, the disinfection process record part stores the compressed results in the distributed file system and uses the Byzantine fault-tolerant consensus algorithm to ensure the consistency and immutability of the data written to the blockchain.

[0013] Furthermore, in the tracing and risk assessment part, when using the graph attention network layer, for each node , its updated features Calculated by the following formula: ;in, Is a node The set of neighbors of ; and The nodes are The common prefix and hash value of is the weight matrix; is the activation function; Is a node With Node The attention coefficient between is calculated as follows: ;in, is a subscript; and The nodes are The common prefix and hash value of It is the vector transpose operation; is a learnable attention vector; and The nodes are The common prefix and hash value of is a ReLU activation function with a negative slope.

[0014] Furthermore, the tracing and risk assessment part uses an abnormal node identification method based on relationship anomalies to define anomaly scores for: ;in, For the updated features; Is a node The hash value of the corresponding medical device.

[0015] Furthermore, in the traceability and risk assessment part, after finding the node most likely to have disinfection anomalies, the zero-knowledge proof protocol is used to obtain relevant information from the blockchain. , using the following steps: The validator randomly chooses a challenge value ; The prover calculates the response value using the following formula ;in Is generating The random salt used when is the hash value of the disinfection process data, is a preset prime number; the verifier checks whether the following equation holds: ; If the equation holds true, it proves the validity of the disinfection record.

[0016] Furthermore, in the traceability and risk assessment part, the traceability of medical devices with disinfection abnormalities is completed through the following steps: obtaining the hash value, timestamp and identifier of the medical device corresponding to the abnormal node from the blockchain; using the identifier of the medical device to locate the specific medical device information in the improved Merkle-Patricia tree structure; decompressing the disinfection records stored in the distributed file system, the decompression process is as follows: ;in, , and They are the inverse operations of the block compression algorithm, moving to the front-end processing, and the arithmetic coding algorithm; , It is The frequency of the symbol.

[0017] The medical device disinfection traceability management system of the present invention has the following beneficial effects: by introducing blockchain technology, the present invention ensures that the disinfection records of medical devices cannot be tampered with during the storage process. The distributed ledger technology of the blockchain can ensure that the disinfection records cannot be maliciously modified or deleted after being written, thereby providing reliable security for the disinfection records. Once the disinfection process record of the medical device is uploaded to the blockchain, a unique hash value will be generated. The hash value is stored in the blockchain as a summary of the disinfection record together with its timestamp and device identifier, ensuring the integrity and transparency of the record. Even if some nodes fail or are attacked, the system can still verify the authenticity of the record through other nodes. The application of blockchain technology makes every change in the disinfection record traceable, greatly improving the transparency and credibility of data management. The present invention ensures that no sensitive information needs to be leaked during the verification process of the disinfection record through a zero-knowledge proof protocol. The application of this protocol enables the system to verify the disinfection record only through the hash value and random challenge value of the disinfection record without directly accessing the specific content of the disinfection record. Such a design greatly improves the privacy protection ability of data, especially in scenarios where the authenticity of data needs to be verified but the data content does not want to be disclosed, zero-knowledge proof can ensure the security and privacy of data. The present invention realizes intelligent detection of disinfection anomalies through a graph attention network and an abnormal node identification method based on relationship anomalies. When the system finds that the disinfection record of a certain medical device may be abnormal, it can analyze the relationship between medical devices through the graph attention network, and identify the nodes with abnormal disinfection records through anomaly score calculation. This method combines the characteristic information of medical devices with the hash value of disinfection data to ensure that the system can quickly and accurately locate equipment with disinfection abnormalities and start the corresponding verification and traceability mechanism. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of the system structure of a medical device disinfection traceability management system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] Example 1: Reference Figure 1, a medical device disinfection traceability management system, the system includes: an initialization part, a disinfection process recording part and a traceability and risk assessment part; the initialization part is used to generate a unique 256-bit identifier containing medical device information for each medical device, regard each medical device as a node, and use an improved Merkle-Patricia tree structure to build a medical device information library based on the identifier corresponding to each node. In the medical device information library, each medical device is a node; the disinfection process recording part is used to use a zero-knowledge proof protocol to encrypt and record the disinfection process data to obtain a disinfection record, use a block sorting compression algorithm to compress the disinfection record, and then compress the compressed The compressed results are stored in a distributed file system, and the hash value, timestamp and corresponding medical device identifier of the compressed results in the distributed file system are written into the blockchain to ensure the consistency and non-tamperability of the data written into the blockchain; the traceability and risk assessment part is used to build a graph neural network for the medical device information library. In the graph neural network, the graph attention network layer is used to capture the relationships between medical devices, and abnormal node identification based on relationship anomalies is performed to find the node most likely to have disinfection anomalies. Then, through the zero-knowledge proof protocol, the hash value, timestamp and corresponding medical device identifier corresponding to the node are obtained from the blockchain to complete the traceability of medical devices with disinfection anomalies.

[0021] Specifically, during the initialization phase, the system generates a unique 256-bit identifier for each medical device. This identifier is not a simple serial number. It contains key information about the medical device, such as device type, manufacturing date, usage history, etc., and uses cryptographic methods to ensure its uniqueness and immutability. In the system, each medical device is regarded as a node, and these nodes are organized into an information library. In order to effectively manage these nodes and their associated information, the system uses an improved Merkle-Patricia tree structure. The Merkle-Patricia tree is an optimized hash tree structure that combines the advantages of the Merkle tree and the Patricia tree. The function of the Merkle tree is to hash the data to ensure the integrity and immutability of the data; the Patricia tree is a compressed dictionary tree structure that can effectively reduce storage space and improve data retrieval efficiency. By combining these two structures, the system can maintain high storage and query efficiency when processing large amounts of medical device data. At the same time, the tree structure of the Merkle-Patricia tree helps to quickly verify the integrity of medical device information and ensure that each node can accurately correspond to its relevant information during the traceability process.

[0022] In the disinfection process record part, the system uses the zero-knowledge proof protocol to encrypt and record the data. Zero-knowledge proof is an advanced cryptographic protocol that allows a party to prove that it knows a secret without revealing the specific content of the secret. In this system, zero-knowledge proof is used for the encrypted record of the disinfection process, ensuring that even if an external party obtains the disinfection data, it cannot decrypt the data. This design effectively protects the privacy security of medical devices, especially in a distributed environment, where data sharing and storage are highly secure. The use of zero-knowledge proof is not only to ensure the security of data encryption, but more importantly, before the data is written to the blockchain, its integrity and authenticity have been verified, thereby avoiding traceability errors caused by incomplete or inaccurate data input. In addition, the system further performs block sorting compression on the disinfection records. The block sorting compression algorithm is a compression algorithm based on data similarity. By sorting and compressing data blocks, the system can effectively reduce the storage space occupied. This is particularly important when facing a large amount of medical device disinfection data, which can significantly improve data storage efficiency and reduce transmission costs. The compressed disinfection records are stored in a distributed file system, and the hash value, timestamp and identifier of the medical device are written to the blockchain. In this way, the system ensures that the sterilization process of each medical device is fully, accurately and tamper-proof recorded.

[0023] Next, in the traceability and risk assessment part, the system uses graph neural network (GNN) to build an association model for the medical device information library. Graph neural network is a deep learning model that can process graph structure data, and is particularly suitable for complex network structures such as medical devices. In this model, each medical device node not only represents its own data, but also establishes associations with other medical device nodes. For example, some medical devices may be used multiple times in the same disinfection environment, and this usage history will affect each other's disinfection status. Through the graph neural network, the system can capture the relationship between these medical devices and establish an overall disinfection information network. In the graph neural network, the system further introduces the graph attention network (GAT), which is a neural network model that can dynamically adjust the weight of the relationship between nodes. Through the graph attention network, the system can automatically identify the nodes with abnormalities in the disinfection process based on the relationship weights between each medical device node and other nodes. The advantage of the graph attention network is that it can automatically optimize the weight distribution between nodes based on actual data. In particular, in the process of identifying disinfection anomalies, the system can quickly focus on the nodes that are most likely to have problems, thereby improving the efficiency and accuracy of identification.

[0024] After identifying the abnormal disinfection node, the system will extract the hash value, timestamp and identifier corresponding to the node from the blockchain through the zero-knowledge proof protocol to further confirm the disinfection record of the medical device. The zero-knowledge proof protocol plays an important role in this process because it ensures that the system can verify the integrity and authenticity of the disinfection process without leaking specific disinfection information. This design not only protects the privacy of the disinfection record, but also ensures that the security and consistency of the data can be guaranteed even when multiple parties verify or query across systems. By obtaining data from the blockchain, the system can effectively trace the abnormal medical devices during the disinfection process and take timely measures to reduce risks. The improved Merkle-Patricia tree structure not only improves the efficiency of medical device information management, but also ensures the integrity and immutability of the data, which is difficult to achieve in the existing technology. At the same time, the introduction of the zero-knowledge proof protocol ensures the security and privacy protection of the data, especially in the encrypted record and data verification process of the disinfection process. This design solves the risk of leakage in traditional encryption methods. Secondly, the block sorting compression algorithm effectively improves the storage efficiency of the system in large-scale medical device management and reduces the consumption of system resources by efficiently compressing the data. The combination of graph neural networks and graph attention networks has greatly improved the intelligence level of the system in identifying abnormal nodes during disinfection. The existing technology usually relies on simple rule matching or linear analysis methods, but this system introduces deep learning models to make the processing of complex relationships between nodes more flexible and accurate. This multi-algorithm fusion design enables the system to achieve efficient traceability management and ensure the security and non-tamperability of data when facing complex medical device disinfection environments. The system's multi-level data verification mechanism and flexible anomaly detection methods provide a comprehensive and reliable solution for medical device disinfection management, and have significant technological progress in the field of medical equipment management.

[0025] Embodiment 2: Based on the previous embodiment, in the initialization part, the medical device information at least includes: device number , the length is bit; date of production , length is 8 bits; production batch , the length is bit; manufacturer code , the length is bit; device type code , the length is All medical device information is spliced ​​in a fixed order to obtain the initial information string ;in, Represents a string concatenation operation; uses a cryptographic hash function to concatenate the initial information string Perform hash calculation to generate a 256-bit hash value As an identifier: ; The hash algorithm has an output length of 256 bits; to ensure that the generated identifier Unique in the entire medical device information database, using the following uniqueness verification mechanism: define the hash set in the medical device information database as ,in is the number of medical devices; Indicates Identifier of a medical device; check the generated Do the following relations satisfy: ; If not satisfied, introduce a random salt value , regenerate a new hash value ; Repeat the above process until a unique .

[0026] Specifically, the process of generating the identifier of a medical device involves multiple key information, such as the device number, production date, production batch, manufacturer code, and device type code. This information is considered to be sufficient basic information to represent the identity of the medical device. During the initialization phase, the system concatenates this information into a string in a fixed order, represented as ,in , , , and They correspond to the device number, production date, production batch, manufacturer code and device type code respectively. Through this splicing operation, the system ensures that the information of each medical device can be recorded completely and standardizedly. Next, the system uses a cryptographic hash function to perform a hash calculation on the spliced ​​string to generate a 256-bit identifier. . The hash function selected is SHA-256, a secure hash algorithm widely used in blockchain technology and cryptography, which can map data of any length to a hash value with a fixed length of 256 bits. The main feature of SHA-256 is its one-way nature, that is, it is easy to generate output from input, but it is almost impossible to deduce input from output. In addition, any slight change in input will result in a huge difference in output, ensuring that the unique identifier of each medical device is highly random and unpredictable. This feature is crucial for the traceability management of medical device disinfection, because it ensures that the identity information of medical devices cannot be tampered with, while ensuring the consistency and security of data.

[0027] In order to ensure the uniqueness of the identifier in the entire system, the system introduces a uniqueness verification mechanism. Since there may be multiple hash identifiers in the medical device information database, the system generates a new identifier. After that, it must verify whether the identifier already exists in the repository. If it already exists, it means that two different medical devices have generated the same identifier in the system. This conflict needs to be resolved by introducing a random salt value. The introduction of is to break the determinism of the hash function, so that even if the initial string is the same , adding different salt values After that, the generated hash value will also be different, thus ensuring uniqueness. This process will be repeated until a hash value different from other identifiers in the library is generated. Through this mechanism, the system ensures the uniqueness of each medical device identifier in the entire information library, thus avoiding potential identification conflicts. The design of the entire identifier generation and verification mechanism fully considers the needs of the medical device disinfection traceability management system, that is, when processing a large amount of medical device data, it must be ensured that each device has a unique and unalterable identity. The choice of SHA-256 hash algorithm can not only effectively ensure data security, but also improve the processing efficiency of the system in large-scale medical device management. By introducing random salt values ​​to deal with identifier conflicts, the system enhances the scalability and robustness when dealing with large-scale medical devices. This design is particularly suitable for the medical field. While ensuring the transparency and traceability of equipment disinfection management, it also guarantees the uniqueness and integrity of equipment information.

[0028] Embodiment 3: Based on the previous embodiment, the initialization part uses the improved Merkle-Patricia tree structure to build the medical device information library, and compares the hash value of the current medical device If there is a common prefix with the hash prefix of the existing nodes in the tree, determine the path position where the node should be inserted based on the longest common prefix; if there is no common prefix, insert a new node and adjust the tree structure so that it still maintains the compression characteristics of the Patricia tree; define each node The storage structure is a triple: ;in, Represents the common prefix from the root to the node; Indicates the hash value of the medical device corresponding to this node ; Represents the list of child nodes of the node; in order to maintain data integrity and verification, the hash value of each node in the improved Merkle-Patricia tree structure is generated by combining the hash values ​​of its child nodes, that is, the hash value of each internal node By the hash value of its child nodes Calculated by the following formula: , For the The hash value of the child nodes, is the total number of child nodes; in order to prevent the improved Merkle-Patricia tree structure from becoming too unbalanced after multiple insertions, affecting query efficiency, a dynamic balancing strategy is introduced to ensure the maximum depth of the improved Merkle-Patricia tree structure by introducing hash value space partitioning and path depth optimization algorithms Controlled in logarithmic time complexity In the range of The following optimization conditions are met: ;in, is the number of nodes in the current tree, is the pre-balancing factor, ; The current depth of the modified Merkle-Patricia tree structure.

[0029] Specifically, when building a medical device information database, the system compares the hash value of the current medical device The longest common prefix is ​​found with the hash prefix of the existing nodes in the tree, and the node insertion path is determined based on this prefix. This prefix matching-based mechanism enables efficient finding of suitable locations when inserting new nodes in the tree structure, thus avoiding unnecessary complex searches. If there is no common prefix, the new node is directly inserted, and the tree structure is adjusted according to the rules of the Patricia tree to keep its compression characteristics. The design of the Patricia tree can effectively reduce the storage of redundant data and improve the efficiency of data retrieval, especially when facing large-scale medical device data, this feature can significantly improve the overall performance of the system. Each node in the system not only stores its own hash value, but also generates the hash value of the parent node through the hash value combination of its child nodes. This recursive hash calculation method ensures the integrity and immutability of the entire tree structure. If any child node in the tree changes, the hash values ​​of all parent nodes related to it will change accordingly, so that any changes in the data can be quickly detected. Such a design provides a highly reliable data verification mechanism for the medical device disinfection traceability management system, ensuring that the disinfection information of each medical device is true and complete during the traceability process.

[0030] However, as more and more medical device data are inserted into the tree, the depth of the tree may continue to increase, resulting in decreased query efficiency. In order to prevent the tree structure from becoming unbalanced, the system introduces a dynamic balancing strategy. This strategy controls the maximum depth of the tree through hash value interval division and path depth optimization algorithm. In the range of logarithmic time complexity. Through this dynamic adjustment, the system ensures that the depth of the tree will not grow excessively even when processing large-scale data, thereby maintaining a low query time complexity. The algorithm for hash value partitioning and path depth optimization is based on the number of nodes in the current tree. and pre-balancing factor To dynamically adjust the depth of the tree. By setting the pre-balancing factor reasonably, the system can automatically adjust the structure of the tree under different data sizes to ensure that the query efficiency is adapted to the data size. Another important feature of the Merkle-Patricia tree is that it can achieve fast data verification. This is particularly critical in the medical device disinfection traceability management system, because the disinfection process of each medical device needs to be recorded and traceable at any time. In the design of the system, any change in any node will cause the hash value of all related parent nodes on the entire path to change. Therefore, if the disinfection information of the medical device is tampered with in the system, the system can detect this change through fast hash verification. This hash-based verification mechanism provides a high degree of security for the system, ensuring that all disinfection records cannot be tampered with.

[0031] Embodiment 4: Based on the previous embodiment, in the disinfection process recording part, the disinfection process data At least include: operator ID, equipment number of medical device, disinfection type, disinfectant name, disinfectant batch, disinfection equipment ID, equipment type, disinfection start time, disinfection end time, disinfection duration, ambient temperature and ambient humidity; map the disinfection process data to an irreversible hash value through a hash function It is a hash algorithm with an output length of 512 bits. The zero-knowledge proof protocol uses Pedersen commitment for encrypted recording. , the generated commitment value for: ; Among them, the commitment value It is a model The value calculated in the group is used for subsequent zero-knowledge proof verification. This commitment value ensures the privacy of the disinfection process data while allowing verification of the integrity and authenticity of the data; is a public parameter, Generators in a group; is a nonlinear term; is a random salt of the commitment value, used in Pedersen commitment to increase randomness to prevent direct deduction through brute force , each time a commitment value is generated, are all different, so that even the same You will also get different commitment values; is a preset prime number used for modular operations; As a disinfection record.

[0032] Specifically, disinfection process data It contains various key information in the disinfection process, including the device number of the medical device, operator ID, disinfectant information, disinfection equipment ID, environmental parameters (such as temperature and humidity), and the specific time and duration of disinfection. These data are crucial for tracing the disinfection process because they can fully record all the details of the medical device from the beginning to the end of disinfection. In this embodiment, in order to protect the privacy of these sensitive data, the system Use the SHA-512 hash function to generate an irreversible hash value . SHA-512 is a widely used cryptographic hash algorithm with an output length of 512 bits. It ensures data security by mapping the input disinfected data to a fixed-length hash value. The irreversibility of the hash value means that even if an attacker obtains the hash value, the original disinfected data cannot be deduced by inversion, thereby effectively preventing data leakage. However, it is not sufficient to protect the privacy of data only through hash functions, because in some scenarios, it is still necessary to verify the authenticity and integrity of the disinfected data without directly exposing the specific disinfected content. To achieve this, this embodiment introduces the Pedersen commitment scheme. Pedersen commitment is a cryptography-based commitment scheme that allows one party to commit to a value without revealing the value, and can verify in a subsequent stage whether the value is consistent with the content originally promised. This scheme is very suitable for use in zero-knowledge proof protocols to ensure the privacy and security of committed data. In the present invention, the hash value of the disinfection process data Used to generate a promise value , this commitment value can be used for subsequent verification without exposing the specific content of the sanitized data.

[0033] The core idea of ​​Pedersen commitment is to pass a random salt to the commitment value. Introducing randomness makes the commitment value generated each time different, even if the disinfection data committed is the same, which can effectively prevent brute force attacks. The calculation formula not only includes The hash value is combined with the commitment value random salt , modular operations, and a nonlinear term. The introduction of nonlinear terms increases the complexity of the calculation and further improves the security of the system. By enhancing the randomness and unpredictability of disinfected data, Pedersen promises not only ensure the privacy of the data, but also confirm the integrity of the data through subsequent verification. Calculate in the group to ensure the promised value It has a high degree of mathematical security. Modular arithmetic is a technique commonly used in cryptography. It can ensure that the calculation results are limited to a specific range, while increasing the complexity of the calculation process to avoid simple derivation or cracking. On this basis, the system introduces generators This design ensures that even if an attacker has the hash value of the disinfected data, , it is still impossible to deduce the disinfection data itself or the commitment value through simple reverse calculation or brute force. In the entire disinfection traceability management system, the introduction of Pedersen commitment provides a powerful mechanism for effective verification operations without disclosing specific disinfection data. When it is necessary to verify whether the disinfection process of a medical device is complete or has been tampered with, the system can verify the commitment value To ensure the consistency and authenticity of the disinfection data. This process ensures that even if the disinfection data is tampered with during storage or transmission, the system can quickly detect the anomaly and refuse to receive the tampered data. In addition, another major advantage of Pedersen's commitment is that it allows the use of zero-knowledge proof protocols. In the disinfection traceability system of the present invention, the introduction of the zero-knowledge proof protocol means that the system can prove the authenticity of the disinfection data to the verification party without revealing the details of the disinfection process. This design is very suitable for scenarios where data privacy needs to be strictly protected, especially in the medical field. The disinfection data of medical devices often involves sensitive information, and its security and privacy must be guaranteed. By using the zero-knowledge proof protocol, the system can not only ensure the integrity and traceability of the disinfection process, but also ensure that this information will not be obtained by unauthorized personnel.

[0034] Embodiment 5: Based on the previous embodiment, the disinfection process record part uses a block sorting compression algorithm to compress the disinfection record. Split into blocks equal to the number of nodes in the modified Merkle-Patricia tree structure ; For each block Apply the block sort compression algorithm: ;in, is to make The position with the smallest lexicographic order among all cyclic shifts; Indicates the block compression algorithm; The resulting application moves to the front-end coding: ;in is the character currently being processed, It’s the Iverson bracket; For the characters processed; Indicates moving to the front-end for processing; The subscript index of the current character; further compression using improved arithmetic coding : ;in, It is The frequency of the symbol; is the arithmetic coding algorithm; the compressed result is: .

[0035] Specifically, the disinfection record data itself contains detailed information such as operator ID, medical device number, disinfection type, disinfection time and environmental conditions. In order to effectively compress and manage this data, the system will Split into several blocks , the size of each block is equal to the number of nodes in the improved Merkle-Patricia tree structure. Through this block mechanism, the system can not only process large-scale data more flexibly, but also ensure that each data block matches the corresponding node in the tree structure, making data management and verification more efficient. After completing the data block, the system The block sorting compression algorithm (BWT) is applied. The core idea of ​​BWT is to generate a new string by circularly shifting the input string and sorting all the shifted strings in lexicographical order. The characteristic of this new string is that the same characters tend to be clustered together to form a more compact form. This feature is very beneficial for subsequent compression algorithms because adjacent repeated characters are easier to be further compressed. In the medical device disinfection traceability management system, this means that the repeated information contained in the disinfection record (such as the same disinfection type, equipment number, etc.) can be effectively clustered together, thereby reducing storage space through the compression algorithm. Specifically, the system performs a compression on each block. When applying BWT, the Calculate all the cyclic shifts and find the cyclic shift position with the smallest lexicographic order This position ensures that the generated new string has the best character clustering effect, which lays the foundation for subsequent encoding and compression. For systems that process a large number of medical device disinfection records, this block sorting algorithm can significantly reduce the redundant information of the original data, allowing the system to use storage space more efficiently when storing a large number of disinfection records.

[0036] After applying the block sorting compression algorithm, the system applies move-to-frontend (MTF) encoding to the BWT results. The principle of MTF encoding is to encode the index position of the current character into an integer value by tracking the order in which the characters appear in the string. Specifically, the system encodes each character according to the character sequence of the BWT result and adjusts it according to its frequency of previous appearance. The advantage of this encoding method is that if a character appears frequently, it will be encoded as a smaller integer, thereby further compressing the data. This process is particularly important for the medical device disinfection traceability management system, because some fields in the disinfection record (such as equipment number, disinfection type, etc.) may appear frequently, and MTF encoding can effectively utilize these repeated patterns, thereby further reducing the storage requirements of the data. Next, the system applies an improved arithmetic coding method to the data after MTF encoding. Arithmetic coding is an encoding method that maps the entire data sequence to a real number interval, and achieves more refined data compression by modeling the frequency of characters. Specifically, the system tracks the frequency of occurrence of each symbol after MTF encoding. , mapping the encoding result of each symbol to a real number interval. The advantage of this method is that it can accurately compress according to the actual distribution of the data, thereby avoiding the redundancy of the traditional fixed-length encoding method. In the medical device disinfection traceability management system, this means that for common fields and symbols in the disinfection record, arithmetic coding can achieve more efficient compression according to its frequency, further reducing the space required for storing and transmitting the disinfection record. The design of Example 5 greatly improves the efficiency of the medical device disinfection traceability management system in processing disinfection records through a combination of a series of compression and encoding algorithms. When faced with a large amount of disinfection data, BWT, MTF and arithmetic coding are combined to significantly reduce the redundant information of the data, so that the system can manage a large number of medical device disinfection records with less storage space. At the same time, these compression algorithms can also ensure that the disinfection data will not lose any key information during the compression process, ensuring the integrity and traceability of the data. In addition, the application of the block sorting compression algorithm enables the system to perform flexible compression optimization according to the characteristics of the data, while MTF and arithmetic coding further improve the accuracy and efficiency of compression. For medical device disinfection records that need to be stored for a long time and frequently accessed, this compression solution can not only significantly save storage costs, but also speed up data transmission and processing, and improve the overall performance of the system. This is of great significance to the large-scale data management needs in the modern medical environment, ensuring that the system can still maintain efficient, safe and reliable operation during the complex disinfection traceability process.

[0037] Embodiment 6: Based on the previous embodiment, after the disinfection process record part stores the compressed results in the distributed file system, the Byzantine fault-tolerant consensus algorithm is used to ensure the consistency and non-tamperability of the data written into the blockchain.

[0038] Specifically, the distributed file system has good scalability and fault tolerance, which enables the system to maintain high efficiency in the storage and management of large-scale medical device data. However, with the increase in data volume and the distributed nature of network nodes, the system must ensure that each node can share and store the same disinfection records, and these records remain tamper-proof in future retrieval and query. Therefore, how to ensure data consistency and security in a multi-node environment becomes a key issue. In order to solve this problem, the system uses the Byzantine fault-tolerant consensus algorithm to ensure that the disinfection records are written into the blockchain, even in the presence of malicious nodes or partial system failures. The basic principle of the Byzantine fault-tolerant algorithm is that in a distributed network, multiple nodes reach a consensus through communication with each other. Even if some nodes may be malicious or fail, the system can still reach a consensus through the majority of nodes to ensure that the data finally written is correct and cannot be tampered with. In the medical device disinfection traceability management system, the application of the Byzantine fault-tolerant consensus algorithm is particularly important. As the core data of the system, disinfection records must be protected from tampering during storage and transmission, especially in a distributed network environment, where there is a risk of node failure or malicious attack. Through the BFT algorithm, the system can effectively resist these risks and ensure that even if some nodes intentionally or unintentionally destroy data, the system can still reach a consensus through the remaining honest nodes to ensure data consistency. Specifically, the process of the Byzantine Fault Tolerant consensus algorithm can be divided into the following steps:

[0039] First, after the compressed results of the disinfection records are stored in the distributed file system, the system will use these compressed records as input and prepare to write them into the blockchain. During this process, each node in the system will verify the data to be written to ensure that the data is consistent with expectations in terms of format, content, and hash value.

[0040] Next, the system will propagate the data to be written among multiple nodes and require each node to return its verification results within a period of time. Once all nodes return the results, the system will summarize and evaluate the results according to the principles of the Byzantine Fault Tolerant consensus algorithm. If most nodes return consistent verification results, the system will consider that the data has reached a consensus and can be safely written into the blockchain. This process ensures that even if a small number of nodes fail or malicious behavior occurs, the system can still reach a consensus through the majority of honest nodes to ensure the correctness and immutability of the data. In addition, the Byzantine Fault Tolerant consensus algorithm also provides high security, ensuring that the system will not be damaged by attacks from a few malicious nodes when processing sensitive disinfection data. This is especially important for the medical device disinfection traceability management system. The disinfection process of medical devices involves the safety and health of patients. Therefore, when tracing these disinfection records, the accuracy and integrity of the data must be ensured, and any form of data tampering may lead to serious consequences. The introduction of the BFT algorithm effectively solves the problem of data tampering that may occur in a distributed environment and provides a strong security guarantee for the system. When the data reaches a consensus through the Byzantine Fault Tolerant consensus algorithm, the system writes the final compressed disinfection record into the blockchain. As a decentralized distributed ledger technology, blockchain has the characteristics of being tamper-proof, ensuring that once data is written, no node can modify it. This design combines the security of blockchain with the fault tolerance of the Byzantine Fault Tolerant consensus algorithm, so that the medical device disinfection traceability management system can still ensure the integrity and consistency of data in the face of malicious attacks or failures.

[0041] Example 7: Based on the previous example, in the tracing and risk assessment part, when using the graph attention network layer, for each node , its updated features Calculated by the following formula: ;in, Is a node The set of neighbors of ; and The nodes are The common prefix and hash value of is the weight matrix; is the activation function; Is a node With Node The attention coefficient between is calculated as follows: ;in, is a subscript; and The nodes are The common prefix and hash value of It is the vector transpose operation; is a learnable attention vector; and The nodes are The common prefix and hash value of is a ReLU activation function with a negative slope.

[0042] Specifically, the core of Example 7 is that a graph attention network (GAT) is introduced into the medical device disinfection traceability management system, which performs feature aggregation and risk assessment on the disinfection data nodes in the system, thereby improving the system's disinfection record management and traceability capabilities in complex scenarios. The graph attention network is a deep learning algorithm based on a graph structure, which is particularly suitable for processing node data with complex associations. In the present invention, the system dynamically weights and aggregates the features of each medical device node through a graph attention network, thereby improving the system's accuracy in identifying disinfection anomalies and relationships between nodes.

[0043] First, in Example 7, the system assigns two core features to each medical device node, namely a common prefix and a hash value. The common prefix is ​​an identifier used to represent the hierarchical relationship between a node and other nodes, while the hash value is an encrypted summary of the medical device disinfection process record. Through these features, the system can distinguish different medical device nodes in the tree structure, while also ensuring the security and non-tamperability of the disinfection process record. In the design of the graph attention network, the feature update of each node depends on the information of its neighboring nodes. Node The neighbor set of Contains other medical device nodes that are directly related to it. The system updates the node by aggregating the features of these neighboring nodes In the feature update process, the system introduces an attention mechanism to calculate the node Each of its neighbor nodes The attention coefficient between , to determine the influence of neighboring nodes on the update of target node features. The calculation of the attention coefficient is based on the similarity of the common prefix and hash value, that is, the more similar the features of two medical device nodes are, the stronger their association in the disinfection process is, and the larger the attention coefficient is. This design enables the system to capture complex relationships between nodes more flexibly, rather than relying solely on a fixed adjacency matrix or a simple graph structure.

[0044] The core function of the attention coefficient is to dynamically adjust the contribution of each neighbor node to the target node. In the medical device disinfection traceability management system, different nodes may have different risks and relevance during the disinfection process. For example, some devices may have a higher relevance because they share the same disinfection equipment or environment, while the connection between other devices may be relatively weak. Through the graph attention network, the system can adaptively weight these node relationships to ensure that during the risk assessment process, the focus is on those neighbor nodes that are highly correlated with the target node. This feature is crucial for the detection of disinfection anomalies because it can help the system identify potential abnormal nodes that may cause safety hazards due to certain specific conditions (such as equipment failure, insufficient disinfection, etc.). In the process of feature aggregation, the system uses the weight matrix The features of neighboring nodes are linearly transformed. The weight matrix is ​​optimized through the training process, which can capture the linear combination of node features and use these combinations to generate more accurate node representations. The system can not only dynamically adjust the relationship between nodes according to the similarity of node features, but also further improve the accuracy of node feature aggregation by optimizing the weight matrix. In the present invention, this feature aggregation method based on the attention mechanism and the weight matrix significantly improves the performance of the system in processing complex disinfection records and medical device networks. The role of the graph attention network in feature updating is to introduce nonlinear transformations to ensure that the system can learn complex feature patterns. Common activation functions such as ReLU or Sigmoid can be used in the feature update part of the graph neural network. Through nonlinear activation, the system can capture the nonlinear relationships implicit in the disinfection records of medical devices, such as the performance differences of certain devices under different disinfection conditions. In practical applications, the activation function helps the system to more accurately assess the potential risks in the disinfection process and helps the system to effectively recognize patterns between different disinfection nodes. A significant advantage of the graph attention network is that it can dynamically adjust the relationship weights between nodes, thereby improving the system's ability to identify disinfection anomalies. In traditional graph neural networks, the relationship between nodes is usually represented by a fixed adjacency matrix, which cannot reflect the dynamic interaction between nodes. The graph attention network uses the attention mechanism to enable the system to adaptively adjust the weights between nodes, especially when dealing with medical device networks with complex interactive relationships. This flexibility is particularly important. By using the graph attention network, the system can more accurately capture the association between disinfection records between medical devices in the traceability and risk assessment parts. Especially when facing a large number of medical devices, the introduction of the graph attention network enables the system to quickly identify disinfection anomalies and potential safety risk nodes in massive data.

[0045] Example 8: Based on the previous example, the tracing and risk assessment part uses an abnormal node identification method based on relationship anomalies to define anomaly scores for: ;in, For the updated features; Is a node The hash value of the corresponding medical device.

[0046] Specifically, the anomaly score It is an important indicator for evaluating whether each medical device node is abnormal. and its neighbor nodes The feature difference between nodes is calculated to measure their similarity. If the feature of a node is significantly different from that of its neighbor nodes, the node is considered a potential abnormal node. The node features are dynamically updated during the system operation. Specifically, the feature It is the result of aggregation and learning through the graph attention network in the previous embodiment. Therefore, the system can effectively capture the behavior patterns of medical devices during the disinfection process through these features. This model introduces feature difference calculation. Specifically, the Euclidean distance is used to measure the node and neighbor nodes The feature difference between them. Euclidean distance The role of is to evaluate the similarity of the features between nodes. The greater the difference, the more obvious the difference in the behavior of the two nodes during the disinfection process, which may mean that the nodes There are abnormal behaviors. For example, if a medical device is not completely sterilized due to an operator error or equipment failure during the sterilization process, its characteristics will be different from other devices sterilized according to the normal process. By calculating this characteristic difference, the system can automatically detect this abnormal situation. The system also takes into account the hash value of the medical device sterilization record Importance in anomaly detection. The hash value is an encrypted summary of the medical device disinfection process, reflecting the disinfection record of each medical device. To ensure the integrity and security of the data, the system introduces the hash value into the calculation of the anomaly score. The hash value not only represents the summary of the disinfection record of each node, but also captures the implicit information in the disinfection data. By combining the hash value with the feature difference, the system can more accurately assess the degree of abnormality of a certain medical device node. Specifically, the exponential function in the formula amplifies the impact of the hash value in a nonlinear way. When the hash value is large, the change in the anomaly score will be more significant. This nonlinear design can help the system more keenly capture anomalies that may be overlooked during the disinfection process.

[0047] In addition, the system adds a normalization factor when calculating the anomaly score ,in is the number of nodes in the network, is a pre-set balancing factor. The role of this normalization factor is to ensure that the system can maintain a reasonable assessment of the anomaly score in networks of different sizes. When the network size is large, the relationship between nodes will become more complex, and the calculation of the anomaly score will become more sophisticated. By logarithmically processing the network size, the system can avoid the problem of decreased anomaly detection due to excessive network size. In other words, even in very complex and large-scale medical device networks, the system can still effectively identify those nodes with significant abnormal behavior. The node identification method based on relationship anomalies brings many advantages. First, it can automatically evaluate the relationship between each node and its neighboring nodes, and accurately determine which nodes have disinfection anomalies through a comprehensive evaluation of feature differences and hash values. This is crucial for the medical device disinfection traceability management system because the accuracy of disinfection records directly affects patient safety. Once a node's disinfection record is significantly different from that of other related nodes, the node will be marked as an anomaly, and the system will immediately conduct risk assessment and subsequent processing on it. Secondly, this method is still efficient and scalable when dealing with large-scale medical device networks. Through feature aggregation and relationship-based anomaly detection in the graph attention network, the system can quickly and comprehensively evaluate a large number of medical devices without sacrificing computational efficiency. This is particularly important for modern hospitals and medical equipment management scenarios, especially when dealing with disinfection records for tens of thousands of medical devices. The system must have strong scalability and real-time response capabilities. Finally, the introduction of hash values ​​not only ensures the immutability of disinfection data, but also plays an important role in anomaly detection. Since the hash value is essentially an encrypted summary of the disinfection process, any tampering with the disinfection data will be directly reflected in the change of the hash value. Therefore, by incorporating the hash value into the calculation of the anomaly score, the system can more comprehensively capture anomalies in the disinfection records and ensure the integrity and security of the data.

[0048] Example 9: Based on the previous example, in the traceability and risk assessment part, after finding the node most likely to have disinfection anomalies, use the zero-knowledge proof protocol to obtain relevant information from the blockchain, assuming that the disinfection record needs to be verified , using the following steps: The validator randomly chooses a challenge value ; The prover calculates the response value using the following formula : ;in Is generating The random salt used when is the hash value of the disinfection process data, is a preset prime number; the verifier checks whether the following equation holds: ; If the equation holds true, it proves the validity of the disinfection record.

[0049] Specifically, in Example 9, when the system detects that the disinfection record of a medical device may be abnormal, it must extract the disinfection record from the distributed storage through the blockchain and verify its legitimacy. Since the blockchain is decentralized and cannot be tampered with, it ensures that the record cannot be maliciously tampered with once it is stored. Through the introduction of the zero-knowledge proof protocol, the system can not disclose the specific data of the disinfection process during the verification process, which greatly enhances the privacy protection ability of the system. The starting point of the verification process is that the verifier randomly generates a challenge value This randomly generated challenge value ensures that each verification is unique and avoids replay attacks. That is, even if an attacker obtains the data from the previous verification process, he cannot reuse the data to cheat. The randomness of increases the security of the system and provides a basis for subsequent calculations. Next, the prover needs to calculate a response value The calculation of this response value relies on the random salt used earlier when generating the sanitized record. , and the hash value of the sanitized data The specific calculation formula is: ,in is a preset prime number that ensures that the result of the operation remains in a finite field. Through this design, the prover can generate a response value that is strongly related to the disinfection record without exposing the disinfection data itself. It is worth noting that the hash value It is an encrypted summary of the disinfection record, reflecting the key information of the entire disinfection process. Any change to the original record will result in a change in the hash value, which ensures that the system can verify the integrity of the data through the hash value.

[0050] The prover generates a response value After that, the verifier needs to verify the validity of the disinfection record through further calculations. The verification process is achieved through a complex formula involving exponential functions and weighted operations. This formula uses a randomly generated challenge value and the response value , combined with disinfection records stored on the blockchain , ensuring that the verifier can confirm the authenticity of the sanitized data without touching it. The left side of the formula contains a series of generators Calculations related to hash values ​​to ensure response values The correctness of It is a calculation based on the disinfection record stored in the blockchain. If the verifier finds that the calculation result on the left matches the right, it can be determined that the disinfection record is valid and has not been tampered with. The cleverness of this verification process is that the verifier does not need to know the specific content of the disinfection record or any details of the data, but only needs to confirm the result obtained by calculating the hash value and the response value. The application of this zero-knowledge proof protocol ensures the privacy of the verification process and avoids the leakage of any medical device disinfection process data. The zero-knowledge proof protocol provides strong privacy protection and data verification capabilities for the medical device disinfection traceability management system. First, it ensures that the verifier cannot obtain the specific content of the disinfection record in the process of confirming the authenticity of the disinfection record. This is crucial for processing sensitive medical data, because the disinfection process records usually involve key information such as detailed information of the equipment, operator data, and environmental conditions. Through this privacy protection mechanism, the system can effectively prevent outsiders or malicious nodes from obtaining sensitive data that should not be disclosed through the verification process. Secondly, the zero-knowledge proof protocol also guarantees the non-tamperability of the disinfection record. Because the hash value Represents the encrypted summary of the disinfection record, and any modification to the disinfection process will be directly reflected in the hash value. Therefore, during the verification process, the system can determine whether the disinfection record has been tampered with by checking the change in the hash value. Combined with the tamper-proof nature of the blockchain, the system can ensure that once the disinfection record is stored in the blockchain, it will not be modified or deleted. This mechanism makes the traceability management of disinfection records more reliable and helps prevent potential fraud or the emergence of erroneous records. In addition, another important feature of the zero-knowledge proof protocol is its efficiency. In Example 9, the verification process can be completed in a relatively short time by calculating random challenge values ​​and response values, while avoiding excessive consumption of computing resources. This is particularly important for the management of disinfection records for large-scale medical equipment, because the number of medical devices is large and the disinfection process is frequent, and the system must be able to complete the verification and traceability operations of a large amount of data in a short time.

[0051] Example 10: Based on the previous example, in the traceability and risk assessment part, the traceability of medical devices with disinfection abnormalities is completed through the following steps: obtaining the hash value, timestamp and identifier of the medical device corresponding to the abnormal node from the blockchain; using the identifier of the medical device to locate the specific medical device information in the improved Merkle-Patricia tree structure; decompressing the disinfection records stored in the distributed file system, the decompression process is as follows: ;in, , and They are the inverse operations of the block compression algorithm, moving to the front-end processing, and the arithmetic coding algorithm; , It is The frequency of the symbol.

[0052] Specifically, the system determines which node has a disinfection anomaly by obtaining the hash value, timestamp, and identifier of the medical device corresponding to the abnormal node from the blockchain. This information cannot be tampered with in the blockchain, ensuring that the system can obtain the most accurate record of the disinfection process. After obtaining the identifier of the medical device, the system uses this identifier to search in the improved Merkle-Patricia tree structure. The Merkle-Patricia tree is an efficient storage and retrieval structure that can quickly locate the information of each medical device. This structure quickly finds a specific device in a large number of medical device records by matching the identifier of each medical device with the node in the tree structure, and further obtains its disinfection record stored in the distributed file system. Next, the system enters the decompression and recovery process of the disinfection record. Since the disinfection record is stored in the distributed file system after compression, a series of inverse operations are required to restore its original content. This process includes three main decompression steps: block sorting algorithm (BWT), move to the front end processing (MTF), and arithmetic coding (AC). First, the system decompresses the compressed data block. Apply the inverse operation of arithmetic coding Arithmetic coding is an efficient coding algorithm based on symbol frequency, which can map the input symbol sequence to a real number interval to achieve data compression. , the system can restore the original symbol sequence from the compressed numerical sequence. Indicates The cumulative frequency of the symbols, frequency information This step ensures that the sanitized record can be accurately decoded based on the frequency of the symbol, thus recovering the original symbol data.

[0053] Next, the system applies the inverse operation of moving to the front-end processing on the arithmetic coded decoded data. Move-to-the-front processing is a coding method used to optimize the processing of repeated symbols. It achieves more efficient coding by moving each occurrence of the symbol to the front of the sequence. , the system is able to restore the original order of the symbols, ensuring that the decompressed data remains consistent with the data before compression. This processing method plays an important role in disinfection records, because some fields in the disinfection process (such as equipment number, disinfectant name, etc.) may be repeated frequently, and MTF encoding can greatly compress these repeated data. Finally, the system applies the inverse operation of the block sorting algorithm (BWT) , restoring the symbol sequence to the original disinfection record data. The BWT algorithm achieves compression by sorting all cyclic shifts of the data so that the same characters are more easily clustered together after encoding. During the decompression process, the inverse operation of BWT can gradually restore the original disinfection record data based on the sorted character sequence. Due to the nature of BWT, it can handle data with similar patterns in disinfection records well, such as multiple medical devices sharing similar disinfection processes or environmental conditions. Through this series of decompression operations, the system successfully restored the complete disinfection record stored in the distributed file system. The recovered records include detailed information on disinfection, such as operator ID, equipment number, disinfection type, disinfectant name, environmental parameters, and the time and duration of disinfection. This information is crucial for further risk assessment and can help the system accurately analyze potential problems in the disinfection process.

[0054] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. Medical equipment disinfection traceability management system, characterized by: The system comprises: an initialization part, a disinfection process recording part and a traceability and risk assessment part; the initialization part is used to generate a unique 256-bit identifier containing medical device information for each medical device, regard each medical device as a node, and use an improved Merkle-Patricia tree structure based on the identifier corresponding to each node to build a medical device information library, in which each medical device is a node; the disinfection process recording part is used to use a zero-knowledge proof protocol to encrypt and record the disinfection process data to obtain a disinfection record, use a block sorting compression algorithm to compress the disinfection record, and then store the compressed result in a distributed file system, and distribute the distributed file system to the system. The hash value, timestamp and identifier of the compressed result in the file system are written into the blockchain to ensure the consistency and non-tamperability of the data written into the blockchain; the traceability and risk assessment part is used to build a graph neural network of the medical device information library. In the graph neural network, the graph attention network layer is used to capture the relationship between medical devices, and abnormal node identification based on relationship anomalies is performed to find the node most likely to have disinfection anomalies. Then, through the zero-knowledge proof protocol, the hash value, timestamp and identifier of the corresponding medical device of the node are obtained from the blockchain to complete the traceability of the medical device with disinfection anomalies; in the initialization part, the medical device information at least includes: equipment number , the length is bit; date of production , length is 8 bits; production batch , the length is bit; manufacturer code , the length is bit; device type code , the length is All medical device information is spliced ​​in a fixed order to obtain the initial information string ;in, Represents a string concatenation operation; uses a cryptographic hash function to concatenate the initial information string Perform hash calculation to generate a 256-bit hash value As an identifier: ; The hash algorithm has an output length of 256 bits; to ensure that the generated identifier Unique in the entire medical device information database, using the following uniqueness verification mechanism: define the hash set in the medical device information database as ,in is the number of medical devices; Indicates Identifier of a medical device; check the generated Do the following relations satisfy: ; If not satisfied, introduce a random salt value , regenerate a new hash value ; Repeat the above process until a unique ; When the initialization part uses the improved Merkle-Patricia tree structure to build the medical device information library, compare the hash value of the current medical device If there is a common prefix with the hash prefix of the existing nodes in the tree, determine the path position where the node should be inserted based on the longest common prefix; if there is no common prefix, insert a new node and adjust the tree structure so that it still maintains the compression characteristics of the Patricia tree; define each node The storage structure is a triple: ;in, Represents the common prefix from the root to the node; Indicates the hash value of the medical device corresponding to this node ; Represents the list of child nodes of the node; in order to maintain data integrity and verification, the hash value of each node in the improved Merkle-Patricia tree structure is generated by combining the hash values ​​of its child nodes, that is, the hash value of each internal node By the hash value of its child nodes Calculated by the following formula: , For the The hash value of the child nodes, is the total number of child nodes; in order to prevent the improved Merkle-Patricia tree structure from becoming too unbalanced after multiple insertions, affecting query efficiency, a dynamic balancing strategy is introduced to ensure the maximum depth of the improved Merkle-Patricia tree structure by introducing hash value space partitioning and path depth optimization algorithms Controlled in logarithmic time complexity In the range of The following optimization conditions are met: ;in, is the number of nodes in the current tree, is the pre-balancing factor, ; The current depth of the modified Merkle-Patricia tree structure.

2. The medical device disinfection traceability management system according to claim 1, characterized in that: In the disinfection process record section, the disinfection process data At least include: operator ID, equipment number of medical device, disinfection type, disinfectant name, disinfectant batch, disinfection equipment ID, equipment type, disinfection start time, disinfection end time, disinfection duration, ambient temperature and ambient humidity; map the disinfection process data to an irreversible hash value through a hash function ; It is a hash algorithm with an output length of 512 bits. The zero-knowledge proof protocol uses Pedersen commitment for encrypted recording. , the generated commitment value for: ； Among them, the commitment value It is a model The value calculated in the group is used for subsequent zero-knowledge proof verification. This commitment value ensures the privacy of the disinfection process data while allowing verification of the integrity and authenticity of the data; is a public parameter, Generators in a group; is a nonlinear term; is a random salt of the commitment value, used in Pedersen commitment to increase randomness to prevent direct deduction through brute force , each time a commitment value is generated, are all different, so that even the same You will also get different commitment values; is a preset prime number used for modular operations; As a disinfection record.

3. The medical device disinfection traceability management system according to claim 2, characterized in that: In the disinfection process record part, when the disinfection record is compressed using the block sorting compression algorithm, the disinfection record Split into blocks equal to the number of nodes in the modified Merkle-Patricia tree structure ; For each block Apply the block sort compression algorithm: ； in, is to make The position with the smallest lexicographic order among all cyclic shifts; Indicates the block compression algorithm; The resulting application moves to the front-end coding: ； in is the character currently being processed, It’s the Iverson bracket; For the characters processed; Indicates moving to the front-end for processing; The subscript index of the current character; further compression using improved arithmetic coding : ； in, , It is The frequency of the symbol; is the arithmetic coding algorithm; the compressed result is: .

4. The medical device disinfection traceability management system according to claim 3, characterized in that: After the disinfection process record part stores the compressed results in the distributed file system, the Byzantine fault-tolerant consensus algorithm is used to ensure the consistency and immutability of the data written to the blockchain.

5. The medical device disinfection traceability management system according to claim 4, characterized in that: In the tracing and risk assessment part, when using the graph attention network layer, for each node , its updated features Calculated by the following formula: ； in, Is a node The set of neighbors of ; and The nodes are The common prefix and hash value of is the weight matrix; is the activation function; Is a node With Node The attention coefficient between is calculated as follows: ； in, is a subscript; and The nodes are The common prefix and hash value of It is the vector transpose operation; is a learnable attention vector; and The nodes are The common prefix and hash value of is a ReLU activation function with a negative slope.

6. The medical device disinfection traceability management system according to claim 5, characterized in that: The tracing and risk assessment part uses an abnormal node identification method based on relationship anomalies to define anomaly scores for: ； in, For the updated features; Is a node The hash value of the corresponding medical device.

7. The medical device disinfection traceability management system according to claim 6, characterized in that: In the traceability and risk assessment part, after finding the node most likely to have disinfection anomalies, the zero-knowledge proof protocol is used to obtain relevant information from the blockchain. If the disinfection record needs to be verified , using the following steps: The validator randomly chooses a challenge value ; The prover calculates the response value using the following formula : ;in Is generating The random salt used when is the hash value of the disinfection process data, is a preset prime number; the verifier checks whether the following equation holds: ； If the equation holds true, the validity of the disinfection record is proven.

8. The medical device disinfection traceability management system according to claim 7, characterized in that: In the traceability and risk assessment part, the traceability of medical devices with disinfection abnormalities is completed through the following steps: obtaining the hash value, timestamp and medical device identifier corresponding to the abnormal node from the blockchain; using the medical device identifier to locate the specific medical device information in the improved Merkle-Patricia tree structure; decompressing the disinfection records stored in the distributed file system. The decompression process is as follows: ； in, , and They are the inverse operations of the block compression algorithm, moving to the front-end processing, and the arithmetic coding algorithm; , It is The frequency of the symbol.

Images