Medical imaging data security management method and system based on blockchain

Abstract

The present invention relates to medical image data security technology, and discloses a medical image data security management method and system based on blockchain, which includes: a server creates a blockchain database, selects an image file, the server generates an image ID for the received image file, calculates the image fingerprint, and stores the blockchain database; uploads the image file to the image cloud database, and uploads the block address to the image cloud record; retrieves the image from the image cloud database or generates an image access record, and stores the blockchain database; updates the blockchain database according to the Merkle tree; and sends the updated content to other hospitals. The present invention stores the image fingerprint and the image access record of the blockchain database, thereby verifying whether the image has changed through the image fingerprint. By storing the image access record, while realizing image data sharing, it ensures that the image data cannot be tampered with and the access record is traceable.

Description

Technical Field

[0001] The present invention relates to medical imaging data security technology, and in particular to a medical imaging data security management method and system based on blockchain. Background Art

[0002] With the continuous development of artificial intelligence technology, artificial intelligence technology has been widely used in medical image auxiliary diagnosis systems. The auxiliary diagnosis system assists diagnosis by intelligently analyzing uploaded medical images. When using the auxiliary diagnosis system, you first need to export the image to be analyzed from the medical image information system (Picture Archiving and Communication Systems, PACS) to the local computer, and then upload it to the auxiliary diagnosis system. In addition, most auxiliary diagnoses are transmitted through the web front end using the upload component or using the local file transfer program. After the file transfer is successful, an AI service analysis request is initiated for analysis.

[0003] While current medical cloud imaging data can be shared, there is a risk of image data leakage or tampering. Existing technologies encrypt image data to address this problem.

[0004] For example, the patent name is: Decentralized management method of medical imaging data based on blockchain, the patent application number is: CN201810746005.7; the application date is: 2018-07-09, the decentralized management method of medical imaging data based on blockchain, the data owner uploads the medical imaging data it has, stores it as an InterPlanetary File System (IPFS) file, takes a hash value for the medical imaging data, and uses the public key of the data owner to encrypt the link list file in IPFS; the data usage right contract uses the private key of the data owner according to the smart contract to decrypt the encrypted link list file in the above steps, and the data user sends the data user public key information and payment tokens to the data usage right contract. When requesting to obtain the medical imaging data, the data usage right contract uses the user public key to re-encrypt the decrypted link list file and sends it to the data user, and generates a blockchain storing the hash address; the data user performs a decryption operation on it through the data user private key to obtain the link list file, and indexes it in the InterPlanetary File System to obtain the actual medical imaging data.

[0005] This technology encrypts the hash address where the data is stored, and uses public and private keys to keep the decryption rights of the address, ensuring that only blockchain users can decrypt the image address and access the image.

[0006] For example, the patent name is a method, device, equipment and medium for transmitting medical image data; the patent application number is CN202010917032.3; the application date is 2020-09-03; the patent records a method, device, equipment and medium for transmitting medical image data, which can first filter out the data in DICOM format, and then further eliminate the illegal information that cannot be used for AI analysis, which not only effectively avoids the impact on AI diagnosis, but also saves bandwidth and realizes the normal transmission of medical image data. The distributed storage technology based on blockchain is used to generate the block information of each sub-file, and the multiple shard data of each sub-file and the corresponding block information of each sub-file are uploaded to the blockchain node, which can not only realize the shard storage of medical image data, but also effectively improve the security of data transmission because the blockchain data cannot be tampered with. In addition, the present invention can also be applied to smart medical care, thereby promoting the construction of smart cities and the development of digital medicine. The present invention also relates to blockchain technology, and the shard data and block information can be stored in the blockchain node.

[0007] This technology stores the entire image file, and uses the tamper-proof nature of blockchain to ensure that the image file is not tampered with.

[0008] However, encrypting the image file or image access address will prolong the access time of the image file, causing the access process to be delayed or wait too long, reducing the efficiency of image sharing. In addition, the above technology cannot guarantee the traceability of image access after the image data is tampered with. Summary of the invention

[0009] In view of the shortcomings of the existing technology that medical cloud imaging data may be tampered with and the access records may be tampered with, the present invention provides a blockchain-based medical imaging data security management method and system. While ensuring that the access speed of the image file is not affected, the integrity of the image file and the traceability of the access record are ensured.

[0010] In order to solve the above technical problems, the present invention is solved by the following technical solutions:

[0011] The medical imaging data security management method based on blockchain includes the following methods:

[0012] Creation of an image blockchain database: Select an image file from any hospital's image management system and transmit it to the server; the server creates a blockchain database for the received image file, which stores image fingerprints, access records, and access record fingerprints;

[0013] Storage of blockchain database: Any hospital's image management system selects an image file and uploads it to the server. The server randomly generates a unique image ID, encodes the image into an image fingerprint through an irreversible hash function, and stores the image fingerprint in a block of the blockchain database.

[0014] When the server accesses an image from the image cloud database, it generates an access record, which is encoded into an access record fingerprint through an irreversible hash function, and the access record and its fingerprint are stored in a block;

[0015] The fingerprint in the block is stored in the structure of a Merkle tree, and the value of its root node (merkleRoot) is used as the address value of the next block; the address of the first block is empty;

[0016] Update of the image blockchain database: The server updates the blockchain database based on the Merkle tree structure;

[0017] Broadcast of the image blockchain database:

[0018] The server broadcasts to the servers of other hospitals and sends update information to the servers of other hospitals.

[0019] By storing the image fingerprints, access records and their fingerprints in the blockchain database, the integrity of the medical imaging data and access records in the region can be effectively protected.

[0020] Preferably, it also includes an integrity judgment of the image file, in which the server compares the image fingerprint in the blockchain database with the image fingerprint calculated in real time for the retrieved image file, thereby making an integrity judgment on the image file.

[0021] By comparing the image fingerprint of the blockchain database, it is ensured that the image data is not tampered with, and the record of user access to the image data is stored.

[0022] As a preferred method, the accuracy of the blockchain database access record is also determined. The server confirms the blockchain information by broadcasting the update of the block, thereby confirming the accuracy of the blockchain database access record. By comparing the information of the blockchain database, it is ensured that the access record of the image data is not tampered with.

[0023] Preferably, the updating step of the image blockchain database includes:

[0024] Step 1: Add a new bottom leaf node to the Merkle tree, and use the image fingerprint of the blockchain database as the hash value of the new bottom leaf node of the Merkle tree;

[0025] Step 2: If the number of leaf nodes in this layer is even after the addition, the hash values ​​of this leaf node and the adjacent isolated leaf nodes are concatenated to form a string, and the hash value calculated by SHA256 replaces the original value of the previous layer;

[0026] If the number of leaf nodes in this layer is odd after the addition, the leaf node will be treated as an isolated leaf node, and its hash value will be calculated by SHA256 to get a new hash value, which will be added to the previous layer;

[0027] Step 3, repeat step 2 for the previous layer;

[0028] Step 4: Repeat step 3 to calculate a unique hash value as the merkleRoot value of the Merkle tree.

[0029] Preferably, the step of determining the integrity of the image file includes:

[0030] The server retrieves the blockchain database ID data from the image cloud recording module and determines the image data retrieval address and the block where the image fingerprint is located;

[0031] The server obtains the image file through the image retrieval address, calculates the hash value of the image file through an irreversible hash function, and performs a Merkle tree proof with the block to which it belongs, thereby confirming the integrity of the blockchain database obtained by the server;

[0032] The server sends the image file to terminal X and generates access records in json format;

[0033] The server uses the current time to concatenate random numbers, generates a hash value as the ID of the access record through an irreversible hash function, adds the access record to the currently unsealed block, and updates the Merkle tree of the block;

[0034] The other server broadcasts the access record and the blockchain database ID to all other servers in region A, and the other servers update the blockchain to complete the synchronization of the blockchain.

[0035] Preferably, the step of determining the accuracy of blockchain database access records includes:

[0036] The user can view all accessed records of the image files through the hospital's in-hospital terminal and confirm the access records;

[0037] The user modifies this access record;

[0038] The user saves the modification, and the front-end server encrypts the modified access record through an irreversible hash function and updates the Merkle tree of this block on the server;

[0039] The hospital’s server detects the update to the Merkle tree in the block and broadcasts the updated access record to all other hospitals’ servers;

[0040] The servers of other hospitals receive the broadcast and locate the block to which the ID belongs;

[0041] If the block is a packaged block, after calculating merkleRoot, it is compared with the preBlockHash of the next block. If the two are different, the blockchain information is wrong, proving that the update is tampered information, cancel the update, and broadcast the error information;

[0042] If the block is not currently packaged, the updated access record will be Merkle-proofed. If the obtained merkleRoot is different from the current value, it is proved that the update is tampered information, the update is canceled, and the error message is broadcast;

[0043] The hospital's server receives the error message broadcast by more than half of the front-end servers, downloads the block from one of the front-end servers, and updates it repeatedly, so the user's tampering fails.

[0044] The blockchain-based medical imaging data security management system includes an imaging cloud database, an imaging cloud record and a server; an imaging file is selected from the imaging management system of any hospital, and the imaging file is transmitted to the server, and the imaging fingerprint is stored in the blockchain database; the imaging file is uploaded to the imaging cloud database, and the imaging file address, imaging fingerprint and imaging ID are recorded in the imaging cloud; the server updates and broadcasts the blockchain database,

[0045] Preferably, an integrity judgment module is also included, and the server performs integrity judgment on the image blockchain database. The server retrieves the image file from the image cloud database and instantly calculates the fingerprint and compares it with the image fingerprint of the blockchain database, thereby performing integrity judgment on the image file.

[0046] Preferably, it also includes an access record accuracy judgment module, and the server judges the accuracy of the access record of the image blockchain database. The server confirms the blockchain information through the user's access record, thereby confirming the accuracy of the blockchain database access record.

[0047] Due to the adoption of the above technical solution, the present invention has significant technical effects: based on the regional medical cloud imaging data of the blockchain, the data is encrypted and verified, and in the process of regional medical imaging cloud data sharing, the data cannot be tampered with and the access records are traceable. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a system composition diagram of the present invention.

[0049] Figure 2 It is a system composition diagram of the present invention.

[0050] Figure 3 It is the Merkle tree graph of the present invention. DETAILED DESCRIPTION

[0051] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The irreversible hash function used in the embodiments is SHA256.

[0052] Example 1

[0053] Region A contains n hospitals. Each hospital deploys a server, which is connected to the hospital's PACS, image cloud database, and image cloud record. The server stores the blockchain database, and the blockchain databases of all servers are kept updated synchronously; the image cloud database stores the impact data of n hospitals in region A; the image cloud record stores the impact data address of these image data and the preBlockHash of the block where the fingerprints of these images are located. The blockchain in this region performs block encapsulation every 10 minutes and creates the next block.

[0054] The medical imaging data security management method based on blockchain includes the following methods:

[0055] Creation of an image blockchain database: Select an image file from any hospital's image management system and transmit it to the server; the server creates a blockchain database for the received image file, which includes the image ID, image address, and image fingerprint;

[0056] The PACS of Hospital 1 archives a CT image file, which is transmitted to its front-end server. The front-end server uses the creation time of the image 2020 / 11 / 20 13:00:14 to concatenate a random number 23, and calculates the hash value as the image file id = 8b9e9050b61cc09a9fd334e83a3b2fdcd65fdc46554cb04f53a823765c5c85d4 through SHA256; the image file is calculated through SHA256 to obtain the hash value as the image fingerprint;

[0057] Storage of the image blockchain database: The server transmits the created blockchain database to the image cloud recording module, and the image cloud recording module stores the received blockchain database;

[0058] Update of the image blockchain database: The server updates the blockchain database based on the Merkle tree;

[0059] Broadcast of the image blockchain database: The server broadcasts to the servers of other hospitals, sending the image ID and image fingerprint to the servers of other hospitals.

[0060] By storing the image ID, image address and image fingerprint in the blockchain database, the security of medical image data in the region can be effectively protected.

[0061] The steps to update the image blockchain database include:

[0062] Step 1: Add a new bottom leaf node to the Merkle tree, and use the image fingerprint of the blockchain database as the hash value of the new bottom leaf node of the Merkle tree;

[0063] Step 2: If the number of leaf nodes in this layer is even after the addition, the hash values ​​of this leaf node and the adjacent isolated leaf nodes are concatenated to form a string, and the hash value calculated by SHA256 replaces the original value of the previous layer;

[0064] If the number of leaf nodes in this layer is odd after the addition, the leaf node will be treated as an isolated leaf node, and its hash value will be calculated by SHA256 to get a new hash value, which will be added to the previous layer;

[0065] Step 3, repeat step 2 for the previous layer;

[0066] Step 4: Repeat step 3 to calculate a unique hash value as the merkleRoot value of the Merkle tree.

[0067] Example 2

[0068] Based on Example 1, this example also includes the integrity judgment of the image file. The server compares the image fingerprint of the blockchain database with the image fingerprint of the cloud recording module to make an integrity judgment on the image file.

[0069] By comparing the image fingerprint of the blockchain database, it is ensured that the image data is not tampered with, and the record of user access to the image data is stored.

[0070] At 15:30:22 on November 20, 2020, user a used the in-hospital terminal X (local area network IP address is 192.168.8.117) to access the image with ID 8b9e9050b61cc09a9fd334e83a3b2fdcd65fdc46554cb04f53a823765c5c85d4 in hospital 2 (public network IP address is ).

[0071] The hospital computer X retrieves the target ID data from the image cloud record through the front-end server, locates the image data retrieval address and the block where the image fingerprint is located;

[0072] After the front-end server obtains the image file through the image retrieval address, it calculates the hash value of the image file through SHA256 and performs Merkle proof with the block to which it belongs to determine whether the data maintains integrity and consistency;

[0073] The front-end server sends the image file to terminal X and generates an access record in json format, the content of which is as follows:

[0074] {

[0075] requestIP:192.168.8.117,

[0076] serverIP:122.225.243.242,

[0077] imageID:8b9e9050b61cc09a9fd334e83a3b2fdcd65fdc46554cb04f53a823765c5c85d4,

[0078] requestTime:2020-11-20 15:30:22

[0079] };

[0080] The front-end server uses the current time: 2020 / 11 / 20 15:31:01 to concatenate the random number 12, and generates the hash value 7ff7858378d31bc63ee7fd4fc98d54de201d2acd0e512be4bbfed2ce2dc7951d as the id of the access record through SHA256, and adds the access record to the current unsealed block, and updates the Merkle tree of the block;

[0081] The front-end server of Hospital 2 broadcasts the access record and ID to all front-end servers in area A, and other front-end servers update the blockchain to complete the synchronization of the blockchain.

[0082] Example 3

[0083] Based on the above embodiment, this embodiment also includes the judgment of the accuracy of the blockchain database access record. The server confirms the blockchain information through the user's access record, thereby confirming the accuracy of the blockchain database access record. By comparing the access record with the information of the blockchain database, it is ensured that the access record of the image data is not tampered with.

[0084] The steps to determine the accuracy of blockchain database access records include:

[0085] On November 21, 2020, at 13:10:10, user b viewed all access records of the image file with id 8b9e9050b61cc09a9fd334e83a3b2fdcd65fdc46554cb04f53a823765c5c85d4 through the in-hospital terminal Y of hospital c (with LAN IP address 192.168.8.117), and located the access record with id 7ff7858378d31bc63ee7fd4fc98d54de201d2acd0e512be4bbfed2ce2dc7951d;

[0086] User B modifies this access record and changes the requestIP to 192.168.8.112;

[0087] User b saves the modification, and the front-end server encrypts the modified access record using SHA256 and updates the Merkle tree of this block on the server.

[0088] The front-end server of hospital c detects the update of the Merkle tree in the block and broadcasts the updated access record to the front-end servers of all other hospitals;

[0089] The front-end servers of other hospitals receive the broadcast and locate the block with the ID 7ff7858378d31bc63ee7fd4fc98d54de201d2acd0e512be4bbfed2ce2dc7951d;

[0090] If the block is a packaged block, the merkleRoot is calculated and compared with the block address of the next block. If the two are different, the blockchain information is wrong, proving that the update is tampered information, the update is canceled, and the error information is broadcast;

[0091] If the block is not currently packaged, the updated access record will be Merkle-proofed. If the obtained merkleRoot is different from the current value, it is proved that the update is tampered information, the update is canceled, and the error message is broadcast;

[0092] The front-end server of hospital c receives the error message broadcast by more than half of the front-end servers, downloads the block from one of the front-end servers, rolls back the update, and user b's tampering fails.

[0093] Example 4

[0094] Based on Example 1, a blockchain-based medical image data security management system includes a cloud recording module and a server; an image file is selected from the image management system of any hospital, and the image file is transmitted to the server; the created blockchain database is transmitted to the image cloud recording module; the server updates and broadcasts the image blockchain database, and uploads the image blockchain database to the cloud recording module, and the cloud recording module stores the received blockchain database.

[0095] The blockchain-based medical imaging data security management system also includes an integrity judgment module. The server makes an integrity judgment on the image blockchain database. The server compares the image fingerprint of the blockchain database with the image fingerprint of the cloud recording module to make an integrity judgment on the image file.

[0096] The blockchain-based medical imaging data security management system also includes an access record accuracy judgment module. The server judges the accuracy of the image blockchain database access record. The server confirms the blockchain information through the user's access record, thereby confirming the accuracy of the blockchain database access record.

Claims

1. A medical imaging data security management method based on blockchain, characterized in that: The following steps are included: Storage of blockchain database: The image management system of any hospital selects an image file and uploads it to the server, which randomly generates a unique image ID. The image is encoded into an image fingerprint through an irreversible hash function and stored in a block in the blockchain database. When the server accesses an image from the image cloud database, it generates an access record, which is encoded into an access record fingerprint through an irreversible hash function, and the access record and its fingerprint are stored in a block; The fingerprint in the block is stored in the structure of a Merkle tree, and the value of its root node, merkleRoot, is used as the address value of the next block; the address of the first block is empty; Storage of image cloud database: The server uploads the image file to the image cloud database for storage; Storage of image cloud records: The server stores the image ID, the address value of the block to which the image fingerprint belongs, and the image cloud database access address of the image file in the image cloud record; Updates to the blockchain database: The blockchain database updates the data in the block based on the Merkle tree structure; Broadcast of the image blockchain database: After the block data is updated, the server broadcasts it to the servers of other hospitals, sending the image ID and image fingerprint to the servers of other hospitals; It also includes the integrity judgment of the image file. The server performs Merkle proof based on the image fingerprint of the blockchain database and the image fingerprint of the image file in the image cloud database, thereby judging the integrity of the image file; It also includes the judgment of the accuracy of the access records of the blockchain database. The blockchain database prevents the access records from being tampered with. When the server finds that the access records in the block have been modified, it updates the Merkle tree and compares it with the corresponding blocks of other servers to determine the accuracy of the access records. The steps to update the image blockchain database include: Step 1: Add a new bottom leaf node to the Merkle tree, and use the new fingerprint as the hash value of the new bottom leaf node; Step 2: If the number of leaf nodes in this layer is even after the addition, the hash values ​​of this leaf node and the adjacent isolated leaf nodes are concatenated to form a string, and the hash value calculated by SHA256 replaces the original value of the previous layer; If the number of leaf nodes in this layer is odd after the addition, the leaf node will be treated as an isolated leaf node, and its hash value will be calculated by SHA256 to get a new hash value, which will be added to the previous layer; Step 3, repeat step 2 for the previous layer; Step 4: Repeat step 3 to calculate a unique hash value as the merkleRoot value of the Merkle tree. The steps for determining the integrity of an image file include: The server retrieves the image cloud database retrieval address of the image file and the block where the image fingerprint is located from the image cloud record according to the image ID; The server obtains the image file from the image cloud database through the image retrieval address, calculates the hash value of the image file through an irreversible hash function, and performs a Merkle tree proof with the block to which it belongs, thereby confirming the integrity of the blockchain database obtained by the server; The server sends the image file to terminal X and generates access records in json format; The server uses the current time to concatenate random numbers, generates a hash value as the ID of the access record through an irreversible hash function, adds the access record to the currently unsealed block, and updates the Merkle tree of the block; Another server broadcasts the access record and blockchain database ID to all other servers in region A, and other front-end servers update the blockchain to complete blockchain synchronization; The steps to determine the accuracy of blockchain database access records include: The user can view all accessed records of the image files through the hospital's in-hospital terminal and confirm the access records; The user modifies this access record; The user saves the modification, and the server generates a fingerprint of the modified access record through an irreversible hash function, and updates the Merkle tree of this block on the server; The hospital’s server detects the update to the Merkle tree in the block and broadcasts the updated access record to all other hospitals’ servers; The servers of other hospitals receive the broadcast and locate the block to which the access record belongs; If the block is a packaged block, after calculating merkleRoot, it is compared with the preBlockHash of the next block. If the two are different, the blockchain information is wrong, proving that the update is tampered information, cancel the update, and broadcast the error information; If the block is not currently packaged, the updated access record will be Merkle-proofed. If the obtained merkleRoot is different from the current value, it is proved that the update is tampered information, the update is canceled, and the error message is broadcast; The hospital's server receives error messages broadcast by more than half of the front-end servers, downloads blocks from one of the servers, and replaces the original blocks, and the user's tampering fails.

2. A blockchain-based medical imaging data security management system, characterized by: Including image cloud records, image cloud database and server; select image files from the image management system of any hospital, upload the image files to the server, generate image ID and image fingerprint, and add the image fingerprint to the blockchain database for storage; upload the image file to the image cloud database, upload the image ID, image file address and image fingerprint to the image cloud record; the server updates and broadcasts the image blockchain database; It also includes an integrity judgment module, where the server retrieves the image file from the image cloud database and calculates the image fingerprint through an irreversible hash function, and compares it with the image fingerprint of the local blockchain database to judge the integrity of the image file; It also includes an access record accuracy judgment module, in which the server judges the accuracy of the access record of the image blockchain database. The server confirms the blockchain information through the user's access record, thereby confirming the accuracy of the blockchain database access record.

Images