Clinical test data management system and method based on block chain technology

Abstract

The invention discloses a clinical test data management system and method based on a block chain technology, and relates to the technical field of data management of block chains. The clinical test data management system based on the block chain technology comprises a clinical test image quality judgment module, an under-chain storage uploading processing module, a permanent storage optimization module and a storage screening processing module. According to the method, the clinical image file is subjected to quality judgment to judge whether to be stored and uploaded under the chain or not, if yes, uploading timeliness processing is carried out on the clinical image file, if not, collection prompting is carried out on the clinical image file, and pin storage optimization is carried out after storage and uploading under the chain are completed; and finally, according to the optimization result, storage screening processing is performed to select the clinical image file stored in the pin, so that the fault tolerance of the storage of the clinical image data under the chain to data loss management is improved, and the problem of data loss caused by the difference of the situation that the clinical image data is permanently stored after being stored and uploaded under the chain in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data management of blockchains, and particularly relates to a clinical trial data management system and method based on blockchain technology. BACKGROUND

[0002] The management of clinical trial data involves a large amount of sensitive information, including patient data, trial processes, drug effects, etc., and the security, transparency and traceability of these information have always been important problems faced by the industry. The management of clinical trial data first formulates a complete data management plan, and designs a data collection tool, wherein the data collection tool design includes designing a CRF table (Case Report Form) and program configuration, the CRF table includes a paper CRF or an EDC (Electronic Data Capture) system electronic CRF (eCRF), then the subject clinical data is acquired through an off-chain acquisition system (such as an electronic data acquisition system EDC, a medical Internet of Things device), and data preprocessing is performed, the main clinical trial data collected includes CRF data, visit information, laboratory data, image data and reports, informed consent, signature, wearing data and physiological indicators, then the free text (such as adverse events, drug names, etc.) is standardized to standard coding, finally, a chain "data freezing" smart contract is enabled at a key node of the trial (such as a data import bank locking period) to prevent data from being tampered with after a key time point, and the data is exported; at the same time, based on chain identity authentication (DID, Decentralized Identity) and permission control contract, the data access range is limited, through off-chain secure storage + on-chain index and permission credentials, trusted data sharing between multiple agencies (such as research institutions, CRO, Contract Research Organization, regulatory agencies) is realized, cross-chain technology (such as Polkadot, Cosmos or Oracle mechanism) is used to connect clinical data between multiple chains, a multi-party collaboration is realized through a consortium chain architecture, and a "on-chain hash + off-chain storage" mode is used to protect data privacy and integrity. All operations are chained to ensure data traceability and tamper resistance.

[0003] For example, the medical test information processing method, device, equipment and readable storage medium disclosed in the patent application with publication number: CN115828304A include: a blockchain network, the blockchain network includes multiple block nodes, and the data contract of each block node stores the identity data corresponding to the target person. After receiving the chain request corresponding to the identity data, the medical test results in the chain request are associated with the identity data and stored in each data contract to update the medical test information of the target person; after receiving the query request corresponding to the identity data, the current latest medical test results of the identity data are obtained from the data contract, and the latest medical test results are sent to the first sending end of the query request.

[0004] For example, the invention patent announcement with announcement number: CN111125755B discloses a medical data processing system and method based on blockchain architecture technology, including: a data acquisition unit, which collects data after the input node responds; a data sending unit, which responds to the connection between units; a collective chain unit, which is used for public distribution of the hash value of the collected data; a private chain unit, which is used to store the hash value of the collected data; a reverse address unit, which is used for distribution of the returned hash value in the collective chain unit to the blockchain private chain module; an interference unit, which is used to form virtual private chain nodes with multiple random addresses, the hash value in the private chain node is randomly generated and the virtual private chain node name is blocked, and the address is the same as the blockchain address. The interference unit responds to at least the private chain hash value generation stage, the hash value return stage, the distribution stage and the block formation stage.

[0005] However, in the process of implementing the technical solutions of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems: In the existing technology, since the imaging data of clinical trial data is usually large in volume, especially high-resolution images, when storage devices (such as hard drives, servers) may malfunction (such as disk damage, server crash), or there is no good naming rules or classification system when the imaging data is uploaded or stored, and the off-chain storage technology does not guarantee a "permanent storage" mechanism, at the same time, during the image data upload process, due to network instability or insufficient bandwidth, the upload may be interrupted or some data may fail to be uploaded. There is a problem of data loss due to differences in the situation of clinical imaging data being permanently stored after being uploaded to the off-chain storage. Summary of the Invention

[0006] The embodiments of the present application provide a clinical trial data management system and method based on blockchain technology, which solves the problem of data loss caused by differences in the situation in which clinical imaging data is permanently stored after being uploaded to the off-chain storage in the prior art, and improves the fault tolerance of the off-chain storage of clinical imaging data to data loss management.

[0007] The embodiment of the application provides a clinical trial data management system based on a blockchain technology, which comprises a clinical trial image quality judgment module, an off-chain storage uploading processing module, a permanent storage optimization module and a storage screening processing module; the clinical trial image quality judgment module is used for performing image quality judgment on an acquired clinical image file, and judging whether the clinical image file is subjected to off-chain storage uploading; the off-chain storage uploading processing module is used for performing uploading timeliness processing on the uploading of the clinical image file if the clinical image file is subjected to off-chain storage uploading, or performing a clinical image file collection prompt; the uploading timeliness processing comprises performing uploading classification to determine the uploading mode of the clinical image file, and performing storage hardware monitoring on the abnormal condition of storage hardware; the permanent storage optimization module is used for acquiring the pin condition of the clinical image file to perform pin storage optimization on the clinical image file after the clinical image file is subjected to off-chain storage uploading; and the storage screening processing module is used for performing storage screening processing based on the result of the pin storage optimization to select the clinical image file subjected to pin storage.

[0008] Further, whether the clinical image file is subjected to off-chain storage uploading is judged, and the specific process is as follows: image quality data of each clinical image in the clinical image file is acquired, the image quality data comprising pixel resolution, contrast, signal-to-noise ratio, filling degree and image compression rate; if all the image quality data is not less than the corresponding reference image quality data, it is indicated that the corresponding clinical image quality is qualified, otherwise a clinical image file collection prompt is performed to prompt a preset worker to reacquire the corresponding clinical image; the reference image quality data comprising a minimum pixel resolution, a minimum contrast, a minimum signal-to-noise ratio, a minimum filling degree and a maximum image compression rate; if the clinical image quality in the clinical image file is qualified, the clinical image file is subjected to off-chain storage uploading, otherwise the clinical image file is not subjected to off-chain storage uploading.

[0009] Further, the uploading of the clinical image file is processed for uploading timeliness, and the specific steps are as follows: A1, obtaining the historical average update frequency and the reference update frequency of the clinical image file of the corresponding category from the preset database; A2, if the historical average update frequency is greater than the reference update frequency, the incremental uploading measure is taken for the clinical image file, and the incremental influence adjustment is performed, otherwise the block uploading measure is taken for the clinical image file, the block uploading means that the clinical image file is uploaded piece by piece after being segmented, and the incremental influence adjustment means that the initial incremental influence factor is compensated based on the difference between the historical average update frequency and the reference update frequency, the initial incremental influence factor includes file bytes and file modification frequency; A3, the storage hardware is monitored for abnormal situation, the corresponding storage hardware influence data is obtained in real time and compared with the reference hardware influence data obtained from the preset database, if the storage hardware influence data is less than the reference hardware influence data, it means that the storage hardware is qualified, and the off-chain storage uploading is performed, otherwise the preset operation and maintenance personnel are notified for hardware operation and maintenance; the storage hardware influence data includes reallocation sector count, pending sector count, non-repairable sector count, addressing error rate and storage disk temperature; the reference hardware influence data includes reallocation sector count limit value, pending sector count limit value, non-repairable sector count limit value, addressing error rate boundary value and storage disk temperature maximum value.

[0010] Further, the pin situation of the clinical image file is obtained to optimize the pin storage of the clinical image file, and the specific process is as follows: obtaining the pin node number of the clinical image file, synchronously obtaining the preset pin node number from the preset database: if the pin node number is less than the preset pin node number, the file pin broadcast is performed, otherwise no additional processing is performed; real-time acquisition of file pin broadcast result, if there is other node applying for storing the corresponding clinical image file, the clinical image file is transmitted to the corresponding block node for pin, otherwise the pin storage optimization is performed.

[0011] Further, the specific steps of pin storage optimization are as follows: the difference between the pin node number and the preset pin node number is denoted as pin node difference; if the pin node difference is not greater than the preset pin node difference, the cache delay time length of the clinical image file is mapped based on the pin node difference in the preset database, and the secondary pin storage optimization is performed on the clinical image file within the cache delay time length, otherwise the first candidate storage score of the clinical image file is mapped based on the pin node difference, and is stored in the first to-be-screened data set which is an empty set initially constructed for storing the first candidate storage score of the clinical image file representing the lowest priority of permanent storage.

[0012] Further, the clinical image file is stored in the second pin storage optimization, and the specific process is as follows: the access frequency of the clinical image file in the cache delay time is obtained, and the reference access frequency is compared: if the access frequency is greater than the reference access frequency, the difference value between the access frequency and the reference access frequency is calculated as the second access difference value, and the cache delay time magnification is obtained by mapping the second access difference value in the preset database; if the access frequency is not greater than the reference access frequency, the second alternative storage score is obtained by mapping the second access difference value in the preset database, and is stored in the second to-be-screened data set constructed in advance, and the second to-be-screened data set represents an empty set in the initial state of the second to-be-screened data set for storing the second alternative storage score representing the priority of the permanent storage of the clinical image file between the lowest priority and the highest priority; the cache delay time is compensated by the cache delay time magnification to obtain the optimized cache delay time, and the clinical image file is stored in the third pin storage optimization in the optimized cache delay time.

[0013] Further, the clinical image file is stored in the second pin storage optimization, and the specific process is as follows: the access frequency of the clinical image file in the cache delay time is obtained, and the reference access frequency is compared: if the access frequency is greater than the reference access frequency, the difference value between the access frequency and the reference access frequency is calculated as the second access difference value, and the cache delay time magnification is obtained by mapping the second access difference value in the preset database; if the access frequency is not greater than the reference access frequency, the second alternative storage score is obtained by mapping the second access difference value in the preset database, and is stored in the second to-be-screened data set constructed in advance, and the second to-be-screened data set represents an empty set in the initial state of the second to-be-screened data set for storing the second alternative storage score representing the priority of the permanent storage of the clinical image file between the lowest priority and the highest priority; the cache delay time is compensated by the cache delay time magnification to obtain the optimized cache delay time, and the clinical image file is stored in the third pin storage optimization in the optimized cache delay time.

[0014] Further, based on the result of the pin storage optimization, the storage screening processing is carried out to select the clinical image file stored in the pin, and the specific steps are as follows: P1, obtaining the remaining storage amount of the block node, and calculating the storage amount of the clinical image file in each to-be-screened data set to obtain the to-be-stored amount; P2, the preset storage ratio is obtained by mapping the to-be-stored amount in the preset database, and the clinical file storage amount is obtained by multiplying the preset storage ratio and the remaining storage amount; P3, based on the clinical file storage amount, the clinical image file is screened in the pin storage according to the order level of each to-be-screened data set; P4, the remaining clinical image file in each to-be-screened data set is processed by cache clearing based on the reverse order level of each to-be-screened data set.

[0015] Further, the specific process of screening the clinical image files for pin storage is as follows: step one, selecting clinical image files in order in the three-level to-be-screened data set for pin storage, and after the remaining clinical file storage capacity cannot store any clinical image file in the three-level to-be-screened data set, step two is executed; step two, selecting clinical image files in order in the two-level to-be-screened data set for pin storage, and after the remaining clinical file storage capacity cannot store any clinical image file in the two-level to-be-screened data set, step three is executed; step three, selecting clinical image files in order in the one-level to-be-screened data set for pin storage, and after the remaining clinical file storage capacity cannot store any clinical image file in the one-level to-be-screened data set.

[0016] The embodiment of the present application provides a clinical trial data management method based on blockchain technology, and the specific steps are as follows: image quality determination is performed on the acquired clinical image files to determine whether to perform off-chain storage and upload of the clinical image files; if off-chain storage and upload is performed, upload timeliness processing of the clinical image files is performed, otherwise, clinical image file collection prompting is performed; the upload timeliness processing includes upload classification to determine the upload mode of the clinical image files, and storage hardware monitoring for abnormal storage hardware conditions; after the off-chain storage and upload of the clinical image files is completed, the pin situation of the clinical image files is acquired to perform pin storage optimization of the clinical image files; storage screening processing is performed based on the results of the pin storage optimization to select the pin-stored clinical image files.

[0017] The one or more technical solutions provided in the embodiment of the present application have at least the following technical effects or advantages: 1. By performing image quality determination on the acquired clinical image files to determine whether to perform off-chain storage and upload of the clinical image files, the quality of the uploaded clinical image files is guaranteed; if off-chain storage and upload is performed, upload timeliness processing of the clinical image files is performed to improve the upload efficiency; otherwise, clinical image file collection prompting is performed; and after the off-chain storage and upload of the clinical image files is completed, the pin situation of the clinical image files is acquired to perform pin storage optimization of the clinical image files, ensuring the efficiency of clinical image file storage; finally, storage screening processing is performed based on the results of the pin storage optimization to select the pin-stored clinical image files, thereby improving the probability of permanent storage of clinical image data to improve the fault tolerance of data loss management of clinical image data off-chain storage, effectively solving the problem of data loss caused by the difference in permanent storage of clinical image data after off-chain storage and upload in the prior art.

[0018] 2、If all the image quality data are not less than the corresponding reference image quality data, it indicates that the corresponding clinical image quality is qualified, otherwise the clinical image file collection prompt is prompted to prompt the preset staff to re-collect the corresponding clinical image, which ensures the quality of the uploaded clinical trial image file. If the clinical image quality in the clinical image file is qualified, the clinical image file is stored and uploaded off-chain, which helps to save storage resources. Otherwise, it is not stored and uploaded off-chain, which not only improves the quality of the clinical image file in the blockchain, but also ensures the effectiveness of the clinical image file stored and uploaded off-chain.

[0019] 3、By comparing the historical average update frequency of the corresponding category of clinical image file with the reference update frequency, if the historical average update frequency is greater than the reference update frequency, the incremental upload measure is taken for the clinical image file and the incremental influence adjustment is performed, otherwise the block upload measure is taken for the clinical image file to improve the efficiency of data upload, and the storage hardware monitoring is performed to ensure the stability of data upload, the corresponding storage hardware influence data is obtained in real time and compared with the reference hardware influence data, if the storage hardware influence data is less than the reference hardware influence data, it indicates that the storage hardware is qualified, and the off-chain storage upload is performed, otherwise the preset operation and maintenance personnel is notified to perform hardware operation and maintenance processing, so as to realize the monitoring of the clinical image file upload to improve the stability of the clinical image file upload.

[0020] 4、By obtaining the remaining storage amount of the block node and counting the storage amount of the clinical image file in each to-be-screened data set to obtain the to-be-stored amount, then the preset storage ratio is obtained by mapping the to-be-stored amount in the preset database, and the clinical file storage amount is obtained by multiplying the preset storage ratio and the remaining storage amount, which is beneficial to more accurately match the node storage capacity, then the clinical image file is screened and pin-stored based on the clinical file storage amount according to the order level of each to-be-screened data set, finally the remaining clinical image file in each to-be-screened data set is cached and cleared based on the reverse order level of each to-be-screened data set, so as to improve the possibility of permanent storage of the clinical image file to prevent data loss in the clinical image file.

[0021] 5. By sequentially selecting clinical image files in the three-level to-be-screened data set for pin storage until the remaining clinical file storage capacity cannot store any clinical image files in the three-level to-be-screened data set, sequentially selecting clinical image files in the two-level to-be-screened data set for pin storage until the remaining clinical file storage capacity cannot store any clinical image files in the two-level to-be-screened data set, sequentially selecting clinical image files in the one-level to-be-screened data set for pin storage until the remaining clinical file storage capacity cannot store any clinical image files in the one-level to-be-screened data set, the overall utilization rate is improved and the waste of fragmented space is reduced, not only the processing efficiency and stability of clinical image storage are improved, but also the fault tolerance of clinical image data chain storage to data loss management is improved. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 A management flowchart of a clinical trial data management system based on a blockchain technology provided by an embodiment of the present application; Figure 2 A structure diagram of a clinical trial data management system based on a blockchain technology provided by an embodiment of the present application; Figure 3 An interface diagram of clinical trial image quality determination in a clinical trial data management system provided by an embodiment of the present application; Figure 4 An interface diagram of permanent storage optimization management in a clinical trial data management system provided by an embodiment of the present application; Figure 5 A flowchart of storage screening processing provided by an embodiment of the present application. DETAILED DESCRIPTION

[0023] The embodiment of the application provides a clinical trial data management system and method based on a blockchain technology, solves the problem of data loss caused by the difference in the permanent storage condition of clinical image data after the off-chain storage and uploading, judges whether the clinical image file is stored and uploaded off-chain by performing image quality determination on the acquired clinical image file, performs uploading timeliness processing on the uploading of the clinical image file if the clinical image file is stored and uploaded off-chain, otherwise, performs a clinical image file collection prompt, acquires the pin condition of the clinical image file after the clinical image file completes the off-chain storage and uploading, performs pin storage optimization on the clinical image file, then obtains the remaining storage amount of a block node and the storage amount of the clinical image file in each to-be-screened data set to obtain a to-be-stored amount, then maps the to-be-stored amount in a preset database to obtain a preset storage ratio, multiplies the preset storage ratio and the remaining storage amount to obtain a clinical file storage amount, then performs pin storage on the clinical image file based on the clinical file storage amount according to the order level of each to-be-screened data set, and finally performs cache clearing processing on the remaining clinical image file in each to-be-screened data set based on the reverse order level of each to-be-screened data set, thereby improving the fault tolerance of the off-chain storage of clinical image data to data loss management.

[0024] The technical solution in the embodiment of the application is used for solving the problem of data loss caused by the difference in the permanent storage condition of clinical image data after the off-chain storage and uploading, and the general idea is as follows: The technical solution in the embodiment of the application is used for solving the problem of data loss caused by the difference in the permanent storage condition of clinical image data after the off-chain storage and uploading, and the general idea is as follows:

[0025] In order to better understand the above technical solution, the above technical solution will be described in detail in combination with the drawings in the specification and specific embodiments.

[0026] As Figure 1As shown in the management flowchart of the clinical trial data management system based on the blockchain technology provided by the embodiment of the present application, the specific logic is as follows: the clinical trial image quality judgment module first judges whether the image quality data is all less than the corresponding reference image quality data, if the image quality data meets the condition, it means that the corresponding clinical image quality is qualified, and enters the off-chain storage uploading processing module, if the image quality data does not meet the condition, the clinical image file collection prompt is carried out, and the image file is prompted to be re-collected; the off-chain storage uploading processing module first carries out uploading classification, if the historical average update frequency is greater than the reference update frequency, the incremental uploading measure is taken for the clinical image file, and the incremental influence adjustment is carried out, if the historical average update frequency is not greater than the reference update frequency, the block uploading measure is taken for the clinical image file, then the storage hardware monitoring is carried out, if the storage hardware influence data is less than the reference hardware influence data, it means that the storage hardware is qualified, and the off-chain storage uploading is carried out, if the storage hardware influence data is greater than the reference hardware influence data, the preset operation and maintenance personnel is informed to carry out hardware operation and maintenance processing, if the image file does not pass the quality judgment, it returns to the clinical trial image quality judgment module for image file collection and re-judgment; in the Pin storage optimization module, if the Pin node number is less than the preset Pin node number, the file Pin broadcast is carried out, otherwise, no additional processing is carried out, the file Pin broadcast result is obtained in real time, if there is other node applying for storing the corresponding clinical image file, the image file is transmitted to the corresponding block node for Pin storage, if there is no other node applying for storage, the Pin storage optimization is carried out; the storage screening processing module first obtains the remaining storage capacity of the block node, and counts the storage capacity of the clinical image file in each to-be-screened data set to obtain the to-be-stored capacity, then the preset storage ratio is obtained by mapping the to-be-stored capacity in the preset database, and the product of the preset storage ratio and the remaining storage capacity is calculated to obtain the clinical file storage capacity, then based on the clinical file storage capacity, the clinical image file is screened for Pin storage according to the order level of each to-be-screened data set, the remaining clinical image file in each to-be-screened data set is processed for cache cleaning based on the reverse order level of each to-be-screened data set, through the above process, not only the possibility of loss of clinical image file is reduced, but also the fault tolerance of off-chain storage of clinical image data to data loss management is improved.

[0027] As a first aspect, as Figure 2 As shown in the structure diagram of the clinical trial data management system based on the blockchain technology provided by the embodiment of the present application, the clinical trial data management system based on the blockchain technology provided by the embodiment of the present application comprises a clinical trial image quality judgment module, an off-chain storage uploading processing module, a permanent storage optimization module and a storage screening processing module.

[0028] The clinical trial image quality assessment module is used to assess the image quality of acquired clinical image files and determine whether to upload them for off-chain storage. Through the tamper-proof nature of the blockchain, the integrity and source credibility of clinical image data are effectively guaranteed.

[0029] The off-chain storage upload processing module is used to perform upload timeliness processing on the clinical image file upload if off-chain storage upload is performed. Otherwise, a clinical image file collection prompt is provided, and upload timeliness processing is performed, including upload classification to determine the upload method of clinical image files, and storage hardware monitoring for storage hardware abnormalities; through upload classification and hardware monitoring mechanism, the timeliness and stability of off-chain storage upload are improved.

[0030] The permanent storage optimization module is used to obtain the PIN status of clinical image files after the off-chain storage upload is completed to optimize the PIN storage of clinical image files; it dynamically optimizes the PIN of image files based on factors such as access frequency and node distribution to reduce storage costs.

[0031] The storage screening processing module is used to perform storage screening processing based on the results of PIN storage optimization to select clinical imaging files stored in PIN; by screening high-priority clinical imaging data according to PIN optimization results, usage frequency and other strategies, the overall storage performance and quality are improved.

[0032] In this embodiment, since the imaging data involved in clinical trials are usually large in size, especially high-resolution images, high requirements are placed on the reliability of the storage system. However, in actual applications, storage devices such as hard drives or servers are subject to failure risks, such as disk damage or system crashes, which may render data unrecoverable. During the upload or storage process of imaging data, if there is a lack of standardized naming rules or classification systems, data management may be chaotic. Existing off-chain storage mechanisms generally do not have "permanent storage" guarantees. Once they fail to be synchronized to the blockchain or long-term storage system, data loss may occur. During the data upload process, due to network fluctuations or insufficient bandwidth, upload interruptions or partial file upload failures may occur. Under the combined effect of the above factors, even if the clinical imaging data is uploaded to the off-chain storage, it may still be at risk of loss due to differences in permanent storage processing. By optimizing the storage of clinical imaging files that have completed off-chain storage optimization upload, not only can the storage efficiency of clinical imaging files be improved, but also the fault tolerance of off-chain storage of clinical imaging data for data loss management can be improved.

[0033] It should be noted that before designing the clinical trial data management system based on blockchain technology, a preset database for storing various system parameters is established by professional preset personnel. This database is used to support the system's automatic processing flow in image data uploading, storage optimization, and screening decisions. The preset database contains but is not limited to the following parameter information: historical average update frequency, reference update frequency, incremental control influence factor, reference hardware performance data, preset pin node quantity, cache delay duration and its amplification ratio, first-level backup storage score value, second-level backup storage score value, third-level backup storage score value, and preset storage ratio, etc. The above parameter values are pre-set by personnel with professional background according to system requirements and hardware resource status. For example, the preset storage ratio is set according to the actual disk storage capacity evaluation and stored in the database by the preset personnel, providing a reference basis for the system when executing the data hierarchical storage strategy.

[0034] Further, whether to store and upload the clinical image file off-chain is determined as follows: First, the image quality data of each clinical image in the clinical image file is obtained, including pixel resolution, contrast, signal-to-noise ratio, fill degree, and image compression rate.

[0035] Among them, the pixel resolution is obtained by reading the fields in the DICOM (Digital Imaging and Communications in Medicine) metadata of the clinical image, the contrast is obtained by calculating the gray standard deviation of the clinical image, the signal-to-noise ratio is obtained by taking the mean of the gray value of the clinical image as the signal and its standard deviation as the noise, the fill degree is obtained by edge detection and counting the proportion of non-zero pixels to the total number of pixels, and the image compression rate is obtained by comparing the byte size of the original (uncompressed) clinical image and the current stored clinical image file.

[0036] Next, the image quality data is compared with the reference image quality data obtained from the preset database: If all the image quality data is not less than the corresponding reference image quality data, it means that the corresponding clinical image quality is qualified, otherwise the clinical image file collection prompt is prompted to prompt the preset staff to re-collect the corresponding clinical image. The reference image quality data is obtained from the preset database and set by the preset staff based on the clinical image quality related data and recorded in the preset database for storage, including the minimum pixel resolution, the minimum contrast, the minimum signal-to-noise ratio, the minimum fill degree, and the maximum image compression rate.

[0037] If the quality of the clinical images in the clinical image file is all qualified, the clinical image file is stored and uploaded off-chain, otherwise it is not stored and uploaded off-chain; through the automatic quality control, the system can automatically determine whether each clinical image meets the quality requirements, avoid uploading low-quality images, and improve the reliability and usability of the data.

[0038] In the embodiment, by systematically evaluating the quality indicators of each image in the clinical image file, the overall quality level of the clinical images in the data management system can be effectively improved, and the fault tolerance of the off-chain storage of the clinical image data in data loss management can be improved. When the image quality is not up to standard, the system can issue a collection prompt in real time to remind the staff to re-collect, improve work efficiency, and reduce manual review costs. At the same time, only the image files with qualified quality are stored and uploaded off-chain, which effectively saves storage resources and avoids redundant and invalid data occupying off-chain storage space. Through pre-quality control, it is ensured that the image data entering the subsequent on-chain management or analysis process has a good imaging foundation, which provides protection for subsequent medical analysis, auxiliary diagnosis and model training.

[0039] As shown in Figure 3 Fig. 1 is an interface schematic diagram of clinical trial image quality determination in a clinical trial data management system provided by the embodiment of the present application. The left side of the interface is an image quality standard setting display, and the right side is a reference image quality data setting area of various image quality data. The administrator has the right to adjust it. The lower middle is the display area of the image quality determination result and the corresponding subsequent operation prompt area, so that the administrator can more intuitively understand the result of the current clinical trial image quality determination and more conveniently adjust the corresponding data.

[0040] Further, the uploading of the clinical image file is subjected to uploading timeliness processing, and the specific steps are as follows: A1, obtaining the historical average update frequency and the reference update frequency of the clinical image file of the corresponding category from the preset database, comparing the historical average update frequency with the reference update frequency to classify the uploading of the clinical image file.

[0041] Specifically, the historical average update frequency represents the mean value of the update frequency recorded in the historical data of the clinical image file of the corresponding category, and the reference update frequency represents the uploading classification reference frequency set by the preset staff. Both of them are calculated or set by the preset staff and then input into the preset database.

[0042] A2, if the historical average update frequency is greater than the reference update frequency, an incremental upload measure is taken for the clinical image file, effectively reducing the transmission bandwidth and time, and an incremental impact adjustment is performed, otherwise a block upload measure is taken for the clinical image file, ensuring data integrity and anti-network jitter, the block upload measure means uploading the clinical image file in blocks after segmentation, the incremental impact adjustment means compensating the initial incremental impact factor based on the difference between the historical average update frequency and the reference update frequency, the initial incremental impact factor includes file bytes and file modification frequency.

[0043] Wherein, the difference between the historical average update frequency and the reference update frequency is the update frequency difference, the update frequency difference is input into the frequency-increment linear regression mapping model to output the corresponding incremental control impact factor, the frequency-increment linear regression mapping model is a linear regression model pre-trained to fit the mapping relationship between the update frequency difference and the incremental control impact factor, the frequency-increment training data is input into the linear regression model, and the frequency-increment linear regression mapping model is obtained by training based on the least squares method through the scikit-learn framework, the frequency-increment training data includes the update frequency difference in the historical time period, and the incremental control impact factor set by the preset staff according to the update frequency difference.

[0044] A3, the storage hardware is monitored for abnormal conditions, the corresponding storage hardware impact data is obtained in real time and compared with the reference hardware impact data obtained from the preset database, if the storage hardware impact data is less than the reference hardware impact data, the storage hardware is qualified, and the off-chain storage upload is performed, otherwise the preset operation and maintenance personnel are notified for hardware operation and maintenance processing, ensuring the stability and security of data upload.

[0045] The storage hardware impact data includes reallocation sector count, pending sector count, unrepairable sector count, addressing error rate and storage disk temperature, the reallocation sector count represents the number of sectors reallocated as bad tracks, the pending sector count represents the number of sectors that have errors when reading the disk but have not been reallocated, the unrepairable sector count represents the number of sectors that cannot be repaired by the hard disk, and the addressing error rate represents the error frequency of the optical head in the addressing process. The reference hardware impact data includes reallocation sector count limit, pending sector count limit, unrepairable sector count limit, addressing error rate threshold and maximum storage disk temperature; in addition, the reference hardware impact data is pre-set by the preset staff and input into the preset database for storage.

[0046] Specifically, the storage hardware impact data is an index related to the health status of the hard disk, which is obtained by accessing the S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) data of the hard disk. The S.M.A.R.T. is a self-monitoring technology built-in modern hard disks and solid state disks.

[0047] In the embodiment, by introducing the incremental control impact factor compensation mechanism based on frequency difference, the upload content can be dynamically adjusted according to the file byte change and the modification frequency, so as to more finely control the upload content, prevent information redundancy or omission, and further improve the upload precision and network load balancing. Once a hardware anomaly is found, the system can immediately trigger an automatic operation and maintenance notification mechanism to remind the operation and maintenance personnel to maintain, reduce business interruption caused by hardware failure, and improve the overall system availability and data security protection capability.

[0048] Further, the pin situation of the clinical image file is acquired to optimize the pin storage of the clinical image file, and the specific process is as follows: In the first step, the pin node number of the clinical image file is acquired, and the preset pin node number is acquired from the preset database. The preset pin node number is set by the preset staff in advance and stored in the preset database. By dynamically monitoring the pin node number of the clinical image file and comparing it with the preset pin node number, if the current pin number is insufficient, a broadcast is actively initiated to encourage more block nodes to participate in file pin storage, which is beneficial to improve the persistence of the clinical image file in the distributed system.

[0049] In the second step, if the pin node number is less than the preset pin node number, the file pin broadcast is performed, otherwise no additional processing is performed.

[0050] In the third step, the file pin broadcast result is acquired in real time. If there is another node applying to store the corresponding clinical image file, the clinical image file is transmitted to the corresponding block node for pinning. Otherwise, the pin storage optimization is performed to avoid resource waste and network congestion, and the "on-demand distribution" is realized.

[0051] In the embodiment, by listening to the pin broadcast result, when other nodes actively apply to store the image file, the system immediately performs file transmission and completes pin solidification, realizes multi-center storage of the file, improves data access efficiency and system response speed. At the same time, if the broadcast does not obtain a responding node, the pin storage optimization process is automatically entered, the fault tolerance of the data loss management of the clinical image data off-chain storage is improved, and the pin optimization scheduling is performed according to the file popularity, historical access frequency or network load state and other factors, the data processing efficiency is improved, and the overall storage cost is reduced.

[0052] As Figure 4 shown, the interface schematic diagram of the permanent storage optimization management in the clinical trial data management system provided by the embodiment of the application is shown. The upper middle part of the interface is a display area of storage optimization progress, which specifically displays the progress from "PIN node number determination", "PIN broadcast", "low-level set", "intermediate set" to "high-level set". The lower middle part is a PIN node setting area, which includes quantitative display of "current PIN node number" and adjustment area of "preset PIN node number". The administrator has the authority to change the "preset PIN node number", which facilitates the administrator to more efficiently manage the permanent storage optimization management module.

[0053] Further, the specific steps of pin storage optimization are as follows: the difference between the pin node number and the preset pin node number is recorded as the pin node difference; if the pin node difference is not greater than the preset pin node difference, the cache delay time length of the clinical image file is mapped based on the pin node difference in the preset database, and the secondary pin storage optimization of the clinical image file is performed within the cache delay time length, otherwise the first alternative storage score is mapped based on the pin node difference of the clinical image file, and is stored in the first to-be-screened data set which is an empty set representing the initial state of the first alternative storage score for storing the clinical image file representing the lowest priority of permanent storage. The first alternative storage score represents the quantitative data of the lowest priority of permanent storage of the corresponding clinical image file.

[0054] It should be noted that the preset pin node number is set by the preset staff and stored in the preset database; and the pin node difference is input into the node difference-first score linear regression mapping model to output the corresponding first alternative storage score. The node difference-first score linear regression mapping model is a linear regression model pre-trained for fitting the mapping relationship between the pin node difference and the first alternative storage score. The node difference-first score training data is input into the linear regression model, and the node difference-first score linear regression mapping model is trained based on the least square method through the scikit-learn framework. The node difference-first score training data includes the pin node difference in the historical time period and the first alternative storage score set by the preset staff according to the pin node difference.

[0055] In the embodiment, by calculating the difference between the current actual pin node number and the preset pin node number, the system can dynamically identify the distribution state of the current file, map different optimization levels according to the difference range, realize more fine-grained storage resource management, and improve the fault tolerance of clinical image data off-chain storage management. Secondly, by generating and dynamically judging the logic of the control parameters stored in the preset database, the system realizes automatic grading optimization and delay processing decision logic without manual intervention, adapts to different medical scenarios and image data sizes, and has good universality and scalability.

[0056] Further, the clinical image file is stored in the second pin for optimization, and the specific process is as follows: the access frequency of the clinical image file in the cache delay time is obtained and compared with the reference access frequency. By comparing the actual access frequency in the cache delay time with the reference access frequency, the system can accurately judge the current heat of the file, and dynamically map the enlarged cache delay time or the second alternative storage score based on this, to realize the intelligent linkage between access behavior and storage strategy.

[0057] In the first case, if the access frequency is greater than the reference access frequency, the difference between the access frequency and the reference access frequency is calculated as the second access difference, and the cache delay time magnification rate is mapped in the preset database based on the second access difference.

[0058] In the second case, if the access frequency is not greater than the reference access frequency, the second alternative storage score is mapped in the preset database based on the second access difference, and stored in the second to-be-screened data set constructed in advance. The second to-be-screened data set represents an initial state of an empty set for storing the second alternative storage score of the clinical image file representing the priority of permanent storage between the lowest priority and the highest priority. The second alternative storage score represents the quantified data of the corresponding clinical image file stored permanently between the lowest priority and the highest priority.

[0059] It should be explained that two mapping tables are constructed in the preset database based on the second access difference. One is the time length magnification rate mapping table, which is a set of mapping relationships between the second access difference and the output cache delay time length magnification rate. When the access frequency is greater than the reference access frequency, the second access difference is input into the time length magnification rate mapping table to output the cache delay time length magnification rate. The second is the second score mapping table, which is a set of mapping relationships between the second access difference and the second alternative storage score. When the access frequency is not greater than the reference access frequency, the second access difference is input into the second score mapping table to output the second alternative storage score.

[0060] The optimized cache delay time length is obtained by compensating the cache delay time length by the cache delay time length amplification rate, and the clinical image file is stored in the three-level pin storage optimization within the optimized cache delay time length.

[0061] In the embodiment, for the clinical image file with the access frequency higher than the reference access frequency, the original cache delay time is compensated by the cache delay time length amplification rate obtained by the second-level access difference mapping, so as to prolong the active period of the hot image file and trigger the three-level pin storage optimization, thereby effectively improving the access efficiency and guarantee level of the popular file. For the file with the access frequency not reaching the reference level, the system maps the second-level access difference into the second-level alternative storage score, and adds the second-level to-be-screened data set, thereby providing clear basis and candidate list for subsequent storage resource recycling, cold and hot layering or archiving, and enhancing the planning and foresight of storage resource use.

[0062] Further, the three-level pin storage optimization is performed on the clinical image file, and the specific process is as follows: within the optimized cache delay time length, the access frequency of the clinical image file is counted and compared with the reference access frequency: On the one hand, if the access frequency is greater than the reference access frequency, it is determined that the corresponding clinical image file is pin stored.

[0063] On the other hand, if the access frequency is not greater than the reference access frequency, a three-level access difference is obtained by the difference between the access frequency and the reference access frequency, a three-level alternative storage score is obtained by mapping the three-level access difference in the preset database, and is stored in the pre-constructed three-level to-be-screened data set, thereby ensuring that the system can reevaluate or layer the non-rigid demand file when the resource is limited. The three-level to-be-screened data set represents a pre-constructed initial state empty set for storing the three-level alternative storage score representing the highest priority of the clinical image file stored permanently. The three-level alternative storage score represents the highest priority of the clinical image file stored permanently.

[0064] It should be noted that the pre-constructed three-level score mapping table is stored in the preset database, and the three-level score mapping table is a set of mapping relationships between the three-level access difference and the three-level alternative storage score. When the access frequency is not greater than the reference access frequency, the obtained three-level access difference is input into the three-level score mapping table, and the corresponding three-level alternative storage score is output.

[0065] In addition, the alternative storage scores in each to-be-screened data set are sorted in descending order, the alternative storage scores are used to quantify the degree of necessity of storage of the corresponding clinical image files pin, the to-be-screened data set includes a tertiary to-be-screened data set, a secondary to-be-screened data set and a primary to-be-screened data set, the alternative storage scores include a tertiary alternative storage score, a secondary alternative storage score and a primary alternative storage score, all the alternative storage scores in the primary, secondary and tertiary to-be-screened data sets are sorted in descending order, clear data priority judgment basis can be provided in system resource adjustment, batch scheduling or long-term archiving strategy, and it is ensured that the image files with the most necessary access are always retained in priority.

[0066] In the embodiment, the tertiary pin optimization strategy judges and maps the dynamic access frequency to reduce data migration and unnecessary cache recovery, optimizes the I / O call link and system response delay, and actively determines and solidifies the hotspot data in the cache life cycle; a closed decision system is formed by the tertiary to-be-screened data set (tertiary, secondary and primary), all data are dynamically circulated and entered into different priority management pools according to the access behavior and mapping score results, redundancy resource waste is avoided, automatic optimization and long-term expansion are supported, and the fault tolerance of the clinical image data chain storage to data loss management is improved.

[0067] As Figure 5As shown, the flowchart of the storage screening process provided by the embodiment of the application is described as follows: the pin node number of the clinical image file is obtained, the preset pin node number is obtained from the preset database synchronously: if the pin node number is less than the preset pin node number, file pin broadcasting is performed, otherwise no additional processing is performed; the file pin broadcasting result is obtained in real time: if there is another node applying for storage of the corresponding clinical image file, the clinical image file is transmitted to the corresponding block node for pin, otherwise pin storage optimization is performed; the difference between the pin node number and the preset pin node number is recorded as the pin node difference value: if the pin node difference value is not greater than the preset pin node difference value, the cache delay time length of the clinical image file is mapped based on the pin node difference value in the preset database, and the secondary pin storage optimization of the clinical image file is performed within the cache delay time length, otherwise the primary alternative storage score is mapped based on the pin node difference value of the clinical image file and stored in the pre-constructed primary screening data set, then the access frequency of the clinical image file within the cache delay time length is obtained and compared with the reference access frequency: if the access frequency is greater than the reference access frequency, the difference between the access frequency and the reference access frequency is calculated as the secondary access difference value, and the cache delay time length magnification is mapped based on the secondary access difference value in the preset database; if the access frequency is not greater than the reference access frequency, the secondary alternative storage score is mapped based on the secondary access difference value in the preset database and stored in the pre-constructed secondary screening data set, the optimized cache delay time length is obtained by compensating the cache delay time length based on the cache delay time length magnification, and the tertiary pin storage optimization of the clinical image file is performed within the optimized cache delay time length. The access frequency of the clinical image file is compared with the reference access frequency within the optimized cache delay time length: if the access frequency is greater than the reference access frequency, the corresponding clinical image file is pin stored, otherwise the difference between the access frequency and the reference access frequency is obtained to obtain the tertiary access difference value, the tertiary alternative storage score is mapped based on the tertiary access difference value in the preset database and stored in the pre-constructed tertiary screening data set. Through the above process, the system can adapt to the resource state in different node environments, which not only helps the system to have good scalability and portability, but also improves the fault tolerance of data loss management of clinical image data off-chain storage.

[0068] Specifically, the storage screening process is performed based on the result of the pin storage optimization to select the pin stored clinical image file, and the specific steps are as follows: P1, obtain the remaining storage amount of the block node, and calculate the storage amount of the clinical image file in each screening data set to obtain the storage amount to be stored; the remaining storage amount is obtained by using the command line tool df (disk free).

[0069] P2, by mapping the to-be-stored amount in the preset database to obtain a preset storage ratio, multiplying the preset storage ratio and the remaining storage amount to obtain a clinical file storage amount, more accurate matching of node storage capability is realized, and system resource utilization is improved.

[0070] It should be noted that the to-be-stored amount is input into the storage linear regression mapping model to output the corresponding preset storage ratio. The storage linear regression mapping model is a linear regression model pre-trained to fit the mapping relationship between the to-be-stored amount and the preset storage ratio. The storage linear regression mapping model is obtained by inputting storage training data into the linear regression model and training based on the least squares method through the scikit-learn framework. The storage training data includes the to-be-stored amount in the historical time period and the preset storage ratio set by the preset staff according to the to-be-stored amount.

[0071] P3, based on the clinical file storage amount, the clinical image files are screened and stored in pin according to the order level of each to-be-screened data set. The order level indicates that each to-be-screened data set is sorted in the order of three-level to-be-screened data set, two-level to-be-screened data set and one-level to-be-screened data set.

[0072] P4, the remaining clinical image files in each to-be-screened data set are cached and deleted based on the reverse order level of each to-be-screened data set. The reverse order level indicates that each to-be-screened data set is sorted in the order of one-level to-be-screened data set, two-level to-be-screened data set and three-level to-be-screened data set.

[0073] In this embodiment, the to-be-stored amount is mapped in the preset database to obtain the corresponding preset storage ratio, and then the actual allocable clinical file storage amount is calculated in combination with the node remaining capacity, so that the system can also prioritize allocation according to the access value under the condition of limited resources, and realize adaptive and efficient pin strategy. At the same time, through the hierarchical priority order (such as three-level -> two-level -> one-level), the files in each data set are screened and stored in pin step by step, so as to ensure that the clinical images with high access frequency and storage value are preferentially obtained for persistent storage, enhance the availability and access performance of key data, and improve the fault tolerance of clinical image data chain storage to data loss management.

[0074] Further, the specific process of screening and pin storing the clinical image files is as follows: Step one, select clinical image files in the three-level to-be-screened data set in order for pin storage, and execute step two after the clinical file storage amount cannot store any clinical image file in the three-level to-be-screened data set, which realizes the priority storage of core image data with high importance and high frequency of use, and improves the practical value of data storage.

[0075] Step two, sequentially select clinical image files in the secondary to-be-screened data set for pin storage until the remaining clinical file storage capacity cannot store any clinical image file in the secondary to-be-screened data set, and then execute step three.

[0076] Step three, sequentially select clinical image files in the primary to-be-screened data set for pin storage until the remaining clinical file storage capacity cannot store any clinical image file in the primary to-be-screened data set; through the hierarchical capacity judgment mechanism, the storage limit is effectively controlled, the overall utilization rate is improved, and the waste of fragmented space is reduced.

[0077] In this embodiment, if one of the clinical image files is larger than the remaining clinical file storage capacity, it is postponed to the next clinical image file; the automatic hierarchical sequential screening and space judgment process can further reduce the risk of manual intervention and error judgment, improve the processing efficiency and stability of clinical image storage, and accurately identify the remaining image files that are not pin-stored under the current node storage capacity. The system can accordingly perform reverse order level cache clearing (primary→secondary→tertiary) to release storage resources and maintain system health running state, further improving the fault tolerance of clinical image data chain storage management for data loss.

[0078] As a second aspect, the application provides a clinical trial data management method based on blockchain technology, and the specific steps are as follows: performing image quality determination on the acquired clinical image files, determining whether to perform off-chain storage and upload of the clinical image files; if the off-chain storage and upload is performed, performing upload timeliness processing on the upload of the clinical image files, otherwise, performing clinical image file collection prompting, upload timeliness processing, including performing upload classification to determine the upload mode of the clinical image files, and performing storage hardware monitoring on abnormal storage hardware conditions. After the off-chain storage and upload of the clinical image files are completed, the pin situation of the clinical image files is acquired to perform pin storage optimization on the clinical image files. Based on the results of the pin storage optimization, storage screening processing is performed to select the pin-stored clinical image files.

[0079] In summary, the embodiment of the present application determines whether to store and upload the acquired clinical image file off-chain by image quality determination of the clinical image file, guarantees the quality of the uploaded clinical image file, if the clinical image file is stored and uploaded off-chain, performs upload timeliness processing on the uploading of the clinical image file to improve the uploading efficiency, otherwise, performs clinical image file acquisition prompting, and after the clinical image file is completed off-chain storage and uploading, acquires the pin situation of the clinical image file to perform pin storage optimization on the clinical image file, ensures the efficiency of clinical image file storage, and finally performs storage screening processing on the pin storage clinical image file based on the result of pin storage optimization, thereby improving the probability of permanent storage of the clinical image data to improve the fault tolerance of data loss management of the off-chain storage of the clinical image data, and effectively solves the problem of data loss caused by the difference in the permanent storage of the clinical image data after off-chain storage and uploading in the prior art.

[0080] Those skilled in the art will appreciate that embodiments of the application can be supplied as a method, a system, or a computer program product. Accordingly, the application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the application can be embodied in the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk memory, CD-ROMs, optical memory, etc.) having computer usable program code embodied therein.

[0081] The present application is described in reference to flowcharts and / or block diagrams of systems, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce an apparatus that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0082] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0083] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0084] Although preferred embodiments of the application have been described herein, it will be apparent to those skilled in the art that various modifications can be made within the scope of the application without departing from the spirit of the application. Accordingly, it is intended that all such possible modifications be included within the scope of the application as defined in the following claims in which the use of the singular is deemed to include the plural, unless otherwise indicated by context.

[0085] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described herein.

Claims

1. Clinical trial data management system based on blockchain technology, characterized by: It includes a clinical trial image quality determination module, an off-chain storage upload processing module, a permanent storage optimization module, and a storage screening processing module; The clinical trial image quality determination module is used to determine the image quality of the acquired clinical image files and determine whether to store and upload the clinical image files off-chain; The off-chain storage upload processing module is used to perform upload timeliness processing on the upload of clinical image files if off-chain storage upload is performed, otherwise, a clinical image file collection prompt is provided. The upload timeliness processing includes performing upload classification to determine the upload method of clinical image files, and performing storage hardware monitoring for storage hardware abnormalities; The permanent storage optimization module is used to obtain the PIN status of the clinical image file after the clinical image file completes the off-chain storage upload to optimize the PIN storage of the clinical image file; The storage screening processing module is used to perform storage screening processing based on the results of PIN storage optimization to select clinical image files stored in PIN.

2. The clinical trial data management system based on blockchain technology according to claim 1, characterized in that: The specific process of determining whether to upload clinical imaging files to off-chain storage is as follows: Acquiring image quality data of each clinical image in the clinical image file, wherein the image quality data includes pixel resolution, contrast, signal-to-noise ratio, fill factor, and image compression rate; If the image quality data are not less than the corresponding reference image quality data, it means that the corresponding clinical image quality is qualified. Otherwise, a clinical image file collection prompt is performed to prompt the preset staff to re-collect the corresponding clinical image. The reference image quality data includes the minimum pixel resolution, minimum contrast, minimum signal-to-noise ratio, minimum fill factor, and maximum image compression rate. If the clinical image quality in the clinical image files is qualified, the clinical image files will be uploaded for off-chain storage, otherwise they will not be uploaded for off-chain storage.

3. The clinical trial data management system based on blockchain technology according to claim 1, characterized in that: The specific steps for processing the timeliness of uploading clinical imaging files are as follows: A1, obtaining the historical average update frequency and reference update frequency of clinical imaging files of the corresponding category from the preset database; A2: If the historical average update frequency is greater than the reference update frequency, incremental uploading measures are adopted for the clinical image files, and incremental impact adjustment is performed. Otherwise, block uploading measures are adopted for the clinical image files. The block uploading means that the clinical image files are segmented and uploaded block by block. The incremental impact adjustment means that the initial incremental impact factors are compensated by the incremental control impact factors obtained by mapping the difference between the historical average update frequency and the reference update frequency in a preset database. The initial incremental impact factors include file bytes and file modification frequency. A3: Perform storage hardware monitoring for storage hardware anomalies, obtain corresponding storage hardware impact data in real time, and compare it with the reference hardware impact data obtained from the preset database. If the storage hardware impact data is less than the reference hardware impact data, it means that the storage hardware is qualified and off-chain storage upload is performed. Otherwise, the preset operation and maintenance personnel are notified to perform hardware operation and maintenance. The storage hardware impact data includes reallocated sector count, pending sector count, unrepairable sector count, addressing error rate, and storage disk temperature; The reference hardware impact data includes a reallocated sector count limit, a pending sector count limit, an unrepairable sector count limit, an addressing error rate threshold, and a maximum storage disk temperature.

4. The clinical trial data management system based on blockchain technology according to claim 1, characterized in that: The PIN status of the clinical image file is obtained to optimize the PIN storage of the clinical image file. The specific process is as follows: Get the pin node number of the clinical image file and simultaneously get the preset pin node number from the preset database: If the number of pin nodes is less than the preset number of pin nodes, the file pin broadcast is performed, otherwise no additional processing is performed; Obtain the file PIN broadcast results in real time. If other nodes apply to store the corresponding clinical image files, the clinical image files will be transferred to the corresponding block node for PIN. Otherwise, PIN storage optimization will be performed.

5. The clinical trial data management system based on blockchain technology as claimed in claim 4, characterized in that: The specific steps of the pin storage optimization are as follows: The difference between the number of pin nodes and the preset number of pin nodes is recorded as the pin node difference; If the pin node difference is not greater than the preset pin node difference, the cache delay duration of the clinical image file is mapped in the preset database based on the pin node difference, and the clinical image file is optimized for secondary pin storage within the cache delay duration. Otherwise, the first-level alternative storage score is obtained based on the pin node difference mapping of the clinical image file and stored in a pre-constructed first-level data set to be screened. The first-level data set to be screened represents a pre-constructed set of first-level alternative storage scores for storing clinical image files with the lowest priority for permanent storage, which is initially empty.

6. The clinical trial data management system based on blockchain technology according to claim 5, characterized in that: The specific process of performing secondary PIN storage optimization on clinical image files is as follows: Get the access frequency of clinical image files within the cache delay period and compare it with the reference access frequency: If the access frequency is greater than the reference access frequency, the difference between the access frequency and the reference access frequency is recorded as the secondary access difference, and the cache delay duration amplification factor is obtained based on the mapping of the secondary access difference in the preset database; If the access frequency is not greater than the reference access frequency, a secondary candidate storage score is obtained by mapping in a preset database based on the secondary access difference, and stored in a pre-constructed secondary to-be-screened data set, wherein the secondary to-be-screened data set represents an initially empty set of secondary candidate storage scores for storing clinical imaging files representing permanent storage priorities between a lowest priority and a highest priority; The cache delay time is compensated by the cache delay time magnification ratio to obtain the optimized cache delay time, and the clinical image files are optimized for three-level PIN storage within the optimized cache delay time.

7. The clinical trial data management system based on blockchain technology according to claim 6, characterized in that: The specific process of performing three-level PIN storage optimization on clinical image files is as follows: During the optimized cache latency period, the access frequency of clinical image files is counted and compared with the reference access frequency: If the access frequency is greater than the reference access frequency, it is determined that the corresponding clinical image file is pin stored; If the access frequency is not greater than the reference access frequency, the difference between the access frequency and the reference access frequency is obtained to obtain a third-level access difference value, and a third-level candidate storage score is obtained by mapping the third-level access difference value in a preset database, and stored in a pre-constructed third-level to-be-screened data set, wherein the third-level to-be-screened data set represents a pre-constructed set of third-level candidate storage scores for storing clinical imaging files representing the highest priority for permanent storage, the initial state of which is empty; The candidate storage scores in each data set to be screened are sorted in reverse order. The candidate storage scores are used to quantify the degree of necessity of storing the corresponding clinical image file pin. The data set to be screened includes a third-level data set to be screened, a second-level data set to be screened, and a first-level data set to be screened. The candidate storage scores include a third-level candidate storage score, a second-level candidate storage score, and a first-level candidate storage score.

8. The clinical trial data management system based on blockchain technology according to claim 1, characterized in that: The storage screening process is performed based on the result of the PIN storage optimization to select clinical image files stored in the PIN. The specific steps are as follows: P1, obtain the remaining storage capacity of the block node, and count the storage capacity of the clinical image files in each data set to be screened to obtain the storage capacity to be stored; P2, by mapping the amount to be stored in the preset database to obtain a preset storage ratio, and multiplying the preset storage ratio by the remaining storage amount to obtain the clinical file storage amount; P3, based on the clinical file storage capacity, the clinical imaging files are screened according to the order level of each dataset to be screened for pin storage; P4: performing cache clearing processing on the remaining clinical image files in each dataset to be screened based on the reverse order of each dataset to be screened.

9. The clinical trial data management system based on blockchain technology according to claim 8, characterized in that: The specific process of screening clinical image files for PIN storage is as follows: Step 1: sequentially select clinical image files from the three-level data set to be screened for pin storage, and execute step 2 after the clinical file storage capacity can no longer store any clinical image files in the three-level data set to be screened; Step 2: sequentially select clinical image files from the secondary dataset to be screened for pin storage, and execute step 3 after the remaining clinical file storage capacity cannot store any clinical image files in the secondary dataset to be screened; Step 3: Select clinical image files in order from the first-level dataset to be screened for pin storage until the remaining clinical file storage capacity can no longer store any clinical image files in the first-level dataset to be screened.

10. A clinical trial data management method based on blockchain technology, characterized by: The specific steps are as follows: Determine the image quality of the acquired clinical imaging files and decide whether to upload them to off-chain storage; If off-chain storage upload is performed, the upload timeliness of clinical image files will be processed; otherwise, a clinical image file collection prompt will be issued. The upload timeliness processing includes upload classification to determine the upload method of clinical image files, and storage hardware monitoring for storage hardware anomalies. After the clinical image file is uploaded to the off-chain storage, the PIN status of the clinical image file is obtained to optimize the PIN storage of the clinical image file; Based on the results of PIN storage optimization, storage screening processing is performed to select clinical image files stored in PIN.

Images