MRD detection information sharing management system

Abstract

The invention relates to the technical field of medical detection, and discloses an MRD detection information sharing management system which comprises a data association fusion module, a hierarchical collection module and a strategy storage module. The fusion module collects original data, sorts correlation and cleans conversion, and establishes a fusion rule base; the grading collection module designs processing grades according to formats, processes and the like, and determines collection grades according to collection frequencies; and the storage module compares the frequency and format information to determine an association level, determines a storage strategy after adjusting a collection level, determines a data storage level based on a block chain technology, adjusts the storage strategy and performs hierarchical storage. Through the multi-module cooperation and block chain technology, structured fusion, hierarchical collection and hierarchical storage of multi-source detection data are achieved, it is guaranteed that the data are real, complete, safe and traceable, and data sharing between medical institutions and standardized research of a detection method are promoted.

Description

Technical Field

[0001] This invention relates to the field of medical testing technology, and more specifically, to an MRD testing information sharing and management system. Background Technology

[0002] In the context of medical informatics and precision medicine, MRD (minimum residual disease) detection, as a core technology for tumor recurrence monitoring and efficacy evaluation, is widely used due to its advantages in multi-dimensional biomarker detection and dynamic monitoring. However, it faces challenges in managing multi-source detection data and sharing it across institutions, resulting in heterogeneous data integration. Currently, it mainly relies on manual data entry or single-machine system storage, making it difficult to achieve the correlation and fusion of detection data from different medical devices, leading to low data utilization efficiency and insufficient security traceability.

[0003] In existing technologies, testing data management is mostly based on storage in a single database, which fails to comprehensively analyze the logical relationships between multiple data sources in the testing process. Furthermore, it lacks dynamic mapping analysis of data acquisition frequency, format structure, and storage level, and has not established a data hierarchical management model. This makes MRD testing data prone to problems such as format incompatibility and incompleteness when shared across institutions. Data authenticity verification and security protection capabilities are weak, severely restricting the standardization research of testing methods and clinical diagnosis and treatment collaboration.

[0004] Therefore, it is necessary to design an MRD detection information sharing and management system. Through the collaboration of data association and fusion module, hierarchical collection module and strategy storage module, and combined with blockchain technology, a data fusion rule base and hierarchical storage mechanism can be established to solve the problems of difficulty in integrating heterogeneous multi-source data, low efficiency of cross-institutional sharing and insufficient security traceability in traditional management systems. Summary of the Invention

[0005] In view of this, the present invention proposes an MRD detection information sharing and management system, which aims to solve the problems of difficulty in integrating heterogeneous multi-source data, low efficiency in cross-institutional sharing, and insufficient security traceability in traditional management systems.

[0006] This invention proposes an MRD detection information sharing and management system, comprising:

[0007] The data association and fusion module is used to collect raw data from various medical devices, sort out the relationships between data according to the detection process logic, establish data association rules, clean and transform the raw data based on the data association rules to form structured multi-source detection data, obtain the format structure information and detection process logic information of each multi-source detection data, and establish a data fusion rule library after matching the format structure of each multi-source detection data with the detection process logic information.

[0008] A data hierarchical collection module is electrically connected to the data association and fusion module. The data hierarchical collection module is used to acquire and receive in real time the acquisition frequency information, the format structure information and detection process logic information of each multi-source detection data from the data association and fusion module, set the data processing level according to the format structure information and detection process logic information, and determine the data collection level according to the acquisition frequency information.

[0009] The data strategy storage module is electrically connected to the data association and fusion module and the data hierarchical collection module, respectively. The data strategy storage module is used to compare the acquisition frequency information and format structure information of each multi-source detection data, determine the data association level according to the comparison result, adjust the data collection level according to the determined data association level and the data processing level to determine the final data collection level, and determine the data storage strategy according to the final data collection level.

[0010] The data sharing module is electrically connected to the data strategy storage module and the data hierarchical collection module, respectively. The data sharing module is used to provide a standardized sharing interface, generate data sharing permission rules based on the storage strategy, and share the multi-source detection data across institutions.

[0011] Furthermore, after the data strategy storage module determines the data storage strategy based on the final data collection level, it also includes:

[0012] The data strategy storage module obtains the type information of each of the multi-source detection data based on blockchain technology, determines the storage level of each of the multi-source detection data according to the obtained type information, adjusts the storage strategy according to the storage level, determines the final storage strategy, and performs hierarchical data storage of each of the multi-source detection data according to the final storage strategy.

[0013] Furthermore, when the data policy storage module adjusts the storage policy according to the storage level, it includes:

[0014] The data strategy storage module obtains the storage level D1 of each multi-source detection data, and presets the standard storage level D0, the first level difference interval [ΔD1, ΔD2], and the second level difference interval [ΔD3, ΔD4].

[0015] Compare storage class D1 with standard storage class D0 to determine whether the storage strategy adjustment has achieved the expected results;

[0016] When D1 = D0, it is determined that the storage strategy adjustment has reached the expected level, and the data strategy storage module performs hierarchical storage according to the current strategy.

[0017] When D1≠D0, calculate the grade difference ΔD=|D1-D0|;

[0018] If ΔD∈[ΔD1,ΔD2], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A1;

[0019] If ΔD∈[ΔD3,ΔD4], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A2;

[0020] Where A1 < A2, and the adjusted storage strategy meets the requirements of data security and access efficiency.

[0021] Furthermore, after adjusting the storage strategy according to the adjustment coefficient, the data strategy storage module further includes:

[0022] The data strategy storage module statistically adjusts the data storage access response time T1, and presets the standard response time T0, the first response difference interval [ΔT1, ΔT2], and the second response difference interval [ΔT3, ΔT4].

[0023] The data strategy storage module compares the access response time T1 with the standard response time T0 to determine whether the storage strategy adjustment has achieved the expected effect.

[0024] When T1≤T0, it is determined that the storage strategy adjustment has achieved the expected effect, and the data strategy storage module maintains the current storage strategy.

[0025] When T1 > T0, calculate the response difference ΔT = |T1 - T0|;

[0026] If ΔT∈[ΔT1,ΔT2], the data strategy storage module triggers the first-level response optimization process;

[0027] If ΔT∈[ΔT3,ΔT4], the data hierarchical collection module redetermines the data collection level.

[0028] Furthermore, after the data classification collection module redetermines the data collection level, it includes:

[0029] The data hierarchical collection module re-acquires the acquisition frequency information F1, format structure information S1, and detection process logic information L1 of each multi-source detection data; and presets the standard acquisition frequency F0, standard format structure features S0, and standard process logic complexity L0.

[0030] The data grading and collection module compares F1 with F0, S1 with S0, and L1 with L0 respectively to determine whether the data collection level meets the expectations.

[0031] When F1 = F0, S1 satisfies the three conditions of S0 and L1 ≤ L0 simultaneously, the data collection level is determined to have met expectations, and the data collection level is determined according to the existing rules.

[0032] When the above conditions are not met, calculate the frequency difference ΔF = |F1-F0|, the format difference ΔS, and the process complexity difference ΔL.

[0033] If any of the indicators ΔF, ΔS, and ΔL is at the lower limit of the corresponding preset difference range, the data hierarchical collection module will adjust the data collection level by one level.

[0034] If any of the indicators ΔF, ΔS, and ΔL is at the upper limit of the corresponding preset difference range, the data hierarchical collection module will make a secondary adjustment to the data collection level.

[0035] Furthermore, after determining the data collection level, the data hierarchical collection module also includes:

[0036] The data hierarchical collection module calculates the adjusted data transmission bandwidth occupancy rate B1, and presets the standard bandwidth occupancy rate B0, the first bandwidth difference interval [ΔB1, ΔB2], and the second bandwidth difference interval [ΔB3, ΔB4].

[0037] Therefore, the data tiered collection module compares the bandwidth utilization rate B1 with the standard bandwidth utilization rate B0 to determine whether the data collection level has achieved the expected results.

[0038] When B1≤B0, it is determined that the data collection level has achieved the expected effect, and the data classification collection module maintains the current collection level.

[0039] When B1 > B0, calculate the bandwidth difference ΔB = |B1 - B0|;

[0040] If ΔB∈[ΔB1,ΔB2], the data association and fusion module optimizes the data format;

[0041] If ΔB∈[ΔB3,ΔB4], the data association and fusion module re-organizes the data association rules.

[0042] Furthermore, after the data association and fusion module re-organizes the data association rules, it includes:

[0043] The data association and fusion module re-analyzes the logical relationships between the original data of each medical device, obtains the current association rule matching degree M1, and presets the standard matching degree M0, the first matching difference interval [ΔM1, ΔM2], and the second matching difference interval [ΔM3, ΔM4].

[0044] The data association and fusion module compares the association rule matching degree M1 with the standard matching degree M0 to determine whether the data association rule sorting has achieved the expected results.

[0045] When M1≥M0, it is determined that the data association rule sorting has reached the expected level, and the data association fusion module performs data cleaning and transformation according to the existing rules;

[0046] When M1 < M0, calculate the matching difference ΔM = |M1 - M0|;

[0047] If ΔM∈[ΔM1,ΔM2], the data association and fusion module adjusts the existing rules according to the optimization coefficient X1;

[0048] If ΔM∈[ΔM3,ΔM4], the data association and fusion module re-establishes the data association rules;

[0049] Where X1 < 1, and the new rule is validated by at least 90% of historical data.

[0050] Furthermore, after the data association and fusion module re-establishes the data association rules, it also includes:

[0051] The data association and fusion module uses new rules to process the test dataset, calculates the integrity C1 and accuracy A1 of the processed data, and presets the standard integrity C0, standard accuracy A0, first integrity difference interval [ΔC1, ΔC2], and first accuracy difference interval [ΔA1, ΔA2].

[0052] The data association and fusion module compares C1 with C0 and A1 with A0 respectively to determine whether the data association rule reconstruction has achieved the expected results.

[0053] When C1≥C0 and A1≥A0, the data association rule reconstruction is deemed to have achieved the expected results, and the new rule is put into formal use.

[0054] When C1 < C0 or A1 < A0, calculate the complete difference ΔC = |C1 - C0| and the accurate difference ΔA = |A1 - A0|:

[0055] If ΔC∈[ΔC1,ΔC2] or ΔA∈[ΔA1,ΔA2], the data association and fusion module optimizes the new rules by combining expert experience;

[0056] If ΔC > [ΔC1, ΔC2] or ΔA > [ΔA1, ΔA2], the data association and fusion module initiates a multi-round iterative optimization process.

[0057] Furthermore, when the data association and fusion module initiates a multi-round iterative optimization process, it includes:

[0058] The data association and fusion module sets the number of iterations N, adjusts the data association rule parameters in each iteration, calculates the comprehensive score R1 of the rule after the iteration, and presets the standard score R0, the first score difference interval [ΔR1,ΔR2], and the second score difference interval [ΔR3,ΔR4].

[0059] The data association and fusion module compares the comprehensive score R1 with the standard score R0 to determine whether the iterative optimization has achieved the expected results.

[0060] When R1≥R0, it is determined that the iterative optimization has reached the expected result, the iteration is terminated and the optimized rule is enabled;

[0061] When R1 < R0, calculate the score difference ΔR = |R1 - R0|;

[0062] If ΔR∈[ΔR1,ΔR2], adjust the rule parameters according to the first step length and continue iterating;

[0063] If ΔR∈[ΔR3,ΔR4], introduce a machine learning algorithm to assist in rule optimization;

[0064] When the number of iterations reaches N and R1 is still less than R0, a manual review process is triggered.

[0065] Furthermore, the manual review process includes: an expert team manually verifying the rules after multiple rounds of iterative optimization, calculating the verification pass rate V1, and setting a standard verification pass rate V0;

[0066] The data association and fusion module compares the verification pass rate V1 with the standard verification pass rate V0 to determine whether the manual review is successful.

[0067] When V1≥V0, the manual review is deemed successful, and the optimized rule is put into official use.

[0068] When V1 < V0, the expert team marks the validation items that fail in the rules and generates a list of modification suggestions;

[0069] The data association and fusion module adjusts the rule parameters according to the modification suggestion list and conducts testing and verification again until the verification pass rate meets the standard or the highest level of review process is triggered.

[0070] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0071] 1. Through the data association and fusion module, the original data from various medical devices is collected, associated, cleaned, and transformed to achieve standardized and structured processing of multi-source testing data, solving the problem of data heterogeneity in traditional systems. This module establishes a data fusion rule base to ensure that data of different formats and testing procedures are accurately matched and integrated, effectively improving data integrity and consistency, and providing a high-quality data foundation for subsequent analysis.

[0072] 2. The data hierarchical collection module sets processing levels based on data format structure and detection process logic, and determines collection levels based on collection frequency, achieving dynamic hierarchical management of data. This mechanism can intelligently adjust collection strategies according to data characteristics, avoiding redundant collection of high-frequency data or omission of low-frequency key data. While improving data processing efficiency, it also optimizes system resource allocation and reduces storage and transmission costs.

[0073] 3. The data strategy storage module uses blockchain technology to obtain data type information and determine storage levels, combining data association levels and processing levels to form a tiered storage strategy. The immutability of blockchain ensures the authenticity of data throughout the entire process from collection to storage, while tiered storage matches storage resources according to data sensitivity and application needs, satisfying data security requirements while improving the clinical query and analysis experience through optimized access efficiency. Attached Figure Description

[0074] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0075] Figure 1 This is a functional block diagram of the MRD detection information sharing and management system provided in an embodiment of the present invention. Detailed Implementation

[0076] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0077] See Figure 1As shown in the figure, this embodiment of the invention proposes an MRD detection information sharing and management system, which includes: a data association and fusion module, a data hierarchical collection module, a data strategy storage module, and a data sharing module.

[0078] Specifically, the data association and fusion module is used to collect raw data from various medical devices, sort out the relationships between data according to the detection process logic, establish data association rules, clean and transform the raw data based on the data association rules to form structured multi-source detection data, obtain the format structure information and detection process logic information of each multi-source detection data, and establish a data fusion rule library after matching the format structure of each multi-source detection data with the detection process logic information.

[0079] The data hierarchical collection module is electrically connected to the data association and fusion module. The data hierarchical collection module is used to acquire and receive the acquisition frequency information, the format structure information and detection process logic information of the data association and fusion module in real time. The data processing level is set according to the format structure information and detection process logic information, and the data collection level is determined according to the acquisition frequency information.

[0080] The data strategy storage module is electrically connected to the data association and fusion module and the data hierarchical collection module, respectively. The data strategy storage module is used to compare the acquisition frequency information and format structure information of each multi-source detection data, determine the data association level based on the comparison results, adjust the data collection level based on the determined data association level and data processing level to determine the final data collection level, and determine the data storage strategy based on the final data collection level.

[0081] The data sharing module is electrically connected to the data strategy storage module and the data hierarchical collection module, respectively. The data sharing module is used to provide a standardized sharing interface, generate data sharing permission rules based on the storage strategy, and share the multi-source detection data across institutions.

[0082] The above embodiments utilize a data association and fusion module to collect, associate, clean, and transform raw data from various medical devices, achieving standardized and structured processing of multi-source testing data and solving the data heterogeneity problem in traditional systems. This module establishes a data fusion rule base to ensure accurate matching and integration of data with different formats and testing procedures, effectively improving data integrity and consistency and providing a high-quality data foundation for subsequent analysis. The data hierarchical collection module sets processing levels based on data format structure and testing procedure logic, and determines collection levels based on collection frequency, achieving dynamic hierarchical data management. This mechanism can intelligently adjust collection strategies according to data characteristics, avoiding redundant collection of high-frequency data or omission of low-frequency key data, improving data processing efficiency while optimizing system resource allocation and reducing storage and transmission costs. The data strategy storage module uses blockchain technology to obtain data type information and determine storage levels, forming a hierarchical storage strategy based on data association levels and processing levels.

[0083] Specifically, after the data strategy storage module determines the data storage strategy based on the final data collection level, it also includes:

[0084] The data strategy storage module uses blockchain technology to obtain the type information of each multi-source detection data, determines the storage level of each multi-source detection data based on the obtained type information, adjusts the storage strategy according to the storage level, determines the final storage strategy, and performs hierarchical data storage of each multi-source detection data according to the final storage strategy.

[0085] The above embodiments demonstrate that the immutability of blockchain ensures the authenticity of data throughout the entire process from collection to storage, while tiered storage matches storage resources according to the sensitivity of the data and application needs, thus satisfying data security requirements and improving the clinical query and analysis experience through access efficiency optimization.

[0086] Specifically, when the data policy storage module adjusts the storage policy according to the storage level, it includes:

[0087] The data strategy storage module obtains the storage level D1 of each multi-source detection data, and presets the standard storage level D0, the first level difference interval [ΔD1, ΔD2], and the second level difference interval [ΔD3, ΔD4].

[0088] Compare storage class D1 with standard storage class D0 to determine whether the storage strategy adjustment has achieved the expected results;

[0089] When D1 = D0, it is determined that the storage strategy adjustment has reached the expected level, and the data strategy storage module performs tiered storage according to the current strategy.

[0090] When D1≠D0, calculate the grade difference ΔD=|D1-D0|;

[0091] If ΔD∈[ΔD1,ΔD2], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A1;

[0092] If ΔD∈[ΔD3,ΔD4], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A2;

[0093] Where A1 < A2, and the adjusted storage strategy meets the requirements of data security and access efficiency.

[0094] Specifically, the standard storage level is set to D0 = 3, the first level difference interval is [ΔD1, ΔD2] = [1, 2], the second level difference interval is [ΔD3, ΔD4] = [3, 4], and the adjustment coefficients are A1 = 0.8 and A2 = 1.2 (A1 < A2). When the storage level of a certain multi-source detection data is D1 = 2, the difference with D0 is ΔD = 1, which belongs to the [1, 2] interval. The storage strategy is adjusted according to A1 = 0.8 (such as fine-tuning the data from the high-speed storage layer to the second-highest speed storage layer and optimizing the backup frequency). If D1 = 6, ΔD = 3, which belongs to the [3, 4] interval, the second strategy is adjusted according to A2 = 1.2 (such as triggering the data compression algorithm upgrade and migrating to the high-density storage layer). The adjusted strategy ensures that the data security access efficiency is not less than 90% of the baseline value through storage resource reallocation.

[0095] The above embodiments achieve fine-grained hierarchical adjustment of storage strategies by dynamically matching the first / second strategy adjustment based on the difference ΔD between storage level D1 and standard storage level D0: when ΔD is in [ΔD1, ΔD2], lightweight strategy optimization is performed through A1, which can accurately adapt to scenarios with slight storage level deviations and avoid resource waste caused by over-adjustment; when ΔD is in [ΔD3, ΔD4], deep strategy reconstruction is implemented through A2, which can effectively cope with the storage resource reallocation needs under severe level deviations, ensuring that different types of multi-source detection data are always matched with the optimal storage strategy in the hierarchical storage architecture, ultimately achieving the technical effect of improving data security access efficiency by 20%-30% and storage resource utilization by 15%-25%.

[0096] Specifically, after adjusting the storage strategy according to the adjustment coefficient, the data strategy storage module also includes:

[0097] The data strategy storage module calculates and adjusts the data storage access response time T1, and sets the preset standard response time T0, the first response difference interval [ΔT1, ΔT2], and the second response difference interval [ΔT3, ΔT4].

[0098] The data strategy storage module compares the access response time T1 with the standard response time T0 to determine whether the storage strategy adjustment has achieved the expected effect.

[0099] When T1≤T0, it is determined that the storage strategy adjustment has achieved the expected effect, and the data strategy storage module maintains the current storage strategy.

[0100] When T1 > T0, calculate the response difference ΔT = |T1 - T0|;

[0101] If ΔT∈[ΔT1,ΔT2], the data strategy storage module triggers the first-level response optimization process;

[0102] If ΔT∈[ΔT3,ΔT4], the data hierarchical collection module redetermines the data collection level.

[0103] Specifically, a standard response time T0 = 100ms is set, with the first response difference interval [ΔT1, ΔT2] = [10ms, 20ms] and the second response difference interval [ΔT3, ΔT4] = [30ms, 40ms]. When the adjusted data storage access response time T1 = 115ms, ΔT = 15ms ∈ [10ms, 20ms], triggering the first-level response optimization process (such as caching strategy adjustment). If T1 = 135ms, ΔT = 35ms ∈ [30ms, 40ms], the data collection level is redefined by the data classification collection module (such as increasing the collection priority of high-frequency data). This mechanism can improve the stability of the system data access response time by 25%-30%, ensuring efficient reading of multi-source detection data.

[0104] The above embodiment monitors the adjusted access response time T1 in real time and compares it with the standard T0. Based on the interval where ΔT is located, it triggers an optimization process to achieve closed-loop verification of the storage strategy adjustment effect: when ΔT∈[ΔT1,ΔT2], the storage strategy details are quickly corrected through the first-level response optimization process; when ΔT∈[ΔT3,ΔT4], the data collection module is linked to reconstruct the data collection logic. This mechanism can improve the stability of data access response time by 25%-30%, and avoid resource consumption caused by over-adjustment through hierarchical processing, ensuring that multi-source detection data always maintains high-efficiency reading performance after the storage strategy is dynamically adjusted.

[0105] Specifically, after the data classification and collection module redefines the data collection level, it includes:

[0106] The data classification and collection module reacquires the acquisition frequency information F1, format structure information S1, and detection process logic information L1 of each multi-source detection data; and presets the standard acquisition frequency F0, standard format structure characteristics S0, and standard process logic complexity L0.

[0107] The data grading and collection module compares F1 with F0, S1 with S0, and L1 with L0 respectively to determine whether the data collection level meets the expectations.

[0108] When F1 = F0, S1 satisfies the three conditions of S0 and L1 ≤ L0 simultaneously, the data collection level is determined to have met expectations, and the data collection level is determined according to the existing rules.

[0109] When the above conditions are not met, calculate the frequency difference ΔF = |F1-F0|, the format difference ΔS, and the process complexity difference ΔL.

[0110] If any of the indicators ΔF, ΔS, and ΔL is at the lower limit of the corresponding preset difference range, the data hierarchical collection module will adjust the data collection level by one level.

[0111] If any of the indicators ΔF, ΔS, and ΔL is at the upper limit of the corresponding preset difference range, the data classification and collection module will adjust the data collection level by two levels.

[0112] Specifically, the standard acquisition frequency is set to F0 = 100Hz, the first frequency difference interval is [ΔF1, ΔF2] = [5Hz, 10Hz], the second frequency difference interval is [ΔF3, ΔF4] = [15Hz, 20Hz], the standard format structure feature is S0 = 1010 (binary), and the standard process logic complexity is L0 = level 3. When the acquisition frequency of a certain multi-source detection data is F1 = 95Hz, ΔF = 5Hz ∈ [5Hz, 10Hz]. Since ΔF is at the lower limit of the corresponding preset difference interval, the data collection level is adjusted by level one (e.g., increasing the acquisition frequency by 10%). If F1 = 80Hz, ΔF = 20Hz ∈ [15Hz, 20Hz]. Since ΔF is at the upper limit of the corresponding preset difference interval, the data collection level is adjusted by level two (e.g., increasing the collection priority level by 2). Through this mechanism, the matching degree between data collection and detection process can be improved by 20%-25%, ensuring the timeliness and completeness of multi-source detection data collection.

[0113] The above embodiments compare the acquisition frequency F1, format structure S1, and process complexity L1 of multi-source detection data with the standard values ​​F0, S0, and L0 in real time. Based on the difference range of ΔF, ΔS, and ΔL, the data collection level is adjusted in stages to achieve dynamic optimization of the data collection strategy: when any indicator is at the lower limit of the difference range, a first-level adjustment (such as fine-tuning the acquisition frequency) is performed, and when it is at the upper limit, a second-level adjustment (such as reconstructing the collection priority) is implemented. This mechanism can improve the matching degree between data collection and detection process by 20%-25%. While ensuring the timeliness and integrity of multi-source data acquisition, it avoids excessive resource consumption through hierarchical processing, ensuring that the efficiency of detection data collection is improved while reducing invalid data transmission by 30%.

[0114] Specifically, after determining the data collection level, the data classification and collection module also includes:

[0115] The data classification and collection module calculates the adjusted data transmission bandwidth occupancy rate B1, and presets the standard bandwidth occupancy rate B0, the first bandwidth difference interval [ΔB1, ΔB2], and the second bandwidth difference interval [ΔB3, ΔB4].

[0116] Therefore, the data tiered collection module compares the bandwidth utilization rate B1 with the standard bandwidth utilization rate B0 to determine whether the data collection level has achieved the expected results.

[0117] When B1≤B0, it is determined that the data collection level has achieved the expected effect, and the data classification collection module maintains the current collection level.

[0118] When B1 > B0, calculate the bandwidth difference ΔB = |B1 - B0|;

[0119] If ΔB∈[ΔB1,ΔB2], the data association and fusion module optimizes the data format;

[0120] If ΔB∈[ΔB3,ΔB4], the data association and fusion module re-organizes the data association rules.

[0121] Specifically, a standard bandwidth utilization rate B0 = 70% is set, with the first bandwidth difference interval [ΔB1, ΔB2] = [5%, 10%] and the second bandwidth difference interval [ΔB3, ΔB4] = [15%, 20%]. When the adjusted data transmission bandwidth utilization rate B1 = 73%, ΔB = 3%, which does not reach the lower limit of the [5%, 10%] interval, so the current collection level is maintained. If B1 = 78%, ΔB = 8% ∈ [5%, 10%], the data association and fusion module starts data format optimization (e.g., using the Zstandard compression algorithm to reduce data volume by 30%). If B1 = 88%, ΔB = 18% ∈ [15%, 20%], then the data association rules are reorganized (e.g., merging high-frequency duplicate data items). This mechanism can reduce the data transmission bandwidth utilization rate by 15%-25%, ensuring the efficiency of multi-source detection data collection while avoiding network congestion.

[0122] The above embodiments achieve intelligent scheduling of network resources by monitoring the data transmission bandwidth utilization rate B1 in real time and comparing it with the standard B0. Based on the interval where ΔB is located, data format optimization or association rule reconstruction is triggered hierarchically. When ΔB∈[ΔB1,ΔB2], the data transmission volume is reduced by optimizing the data format (such as enabling the Zstandard compression algorithm). When ΔB∈[ΔB3,ΔB4], high-frequency data items are merged by reorganizing the association rules. This mechanism can reduce the data transmission bandwidth utilization rate by 15%-25%. While ensuring a 20% improvement in the efficiency of multi-source detection data collection, it avoids data transmission delays caused by bandwidth congestion, and achieves coordinated optimization of network resource utilization and data transmission efficiency during real-time sharing of detection data.

[0123] Specifically, after the data association and fusion module reorganized the data association rules, it included:

[0124] The data association and fusion module re-analyzes the logical relationships between the original data of each medical device, obtains the current association rule matching degree M1, and presets the standard matching degree M0, the first matching difference interval [ΔM1, ΔM2], and the second matching difference interval [ΔM3, ΔM4].

[0125] The data association and fusion module compares the matching degree of the association rule M1 with the standard matching degree M0 to determine whether the data association rule sorting has met expectations.

[0126] When M1≥M0, it is determined that the data association rule sorting has reached the expected level, and the data association fusion module performs data cleaning and transformation according to the existing rules;

[0127] When M1 < M0, calculate the matching difference ΔM = |M1 - M0|;

[0128] If ΔM∈[ΔM1,ΔM2], the data association and fusion module adjusts the existing rules according to the optimization coefficient X1;

[0129] If ΔM∈[ΔM3,ΔM4], the data association and fusion module re-establishes the data association rules;

[0130] Where X1 < 1, and the new rule is validated by at least 90% of historical data.

[0131] Specifically, the standard matching degree M0 is set to 85%, the first matching difference interval [ΔM1, ΔM2] = [5%, 10%], the second matching difference interval [ΔM3, ΔM4] = [15%, 20%], and the optimization coefficient X1 = 0.9 (X1 < 1). When the current association rule matching degree M1 = 80%, ΔM = 5% ∈ [5%, 10%], and the existing rule is adjusted according to X1 = 0.9 (such as fine-tuning the association weight of data fields). If M1 = 65%, ΔM = 20% ∈ [15%, 20%], then the data association rule is re-established, and the new rule needs to be verified by 90% of historical data (such as testing 100 sets of historical detection data, at least 90 sets meet the rule matching requirements). Through this mechanism, the matching degree of data association rule can be improved by 15%-25%, ensuring the accuracy and consistency of multi-source detection data cleaning and transformation.

[0132] The above embodiments achieve dynamic and precise adjustment of data association rules through a hierarchical optimization mechanism based on the difference ΔM between the matching degree M1 and the standard M0 of the association rule: when ΔM∈[ΔM1,ΔM2], the existing rules are optimized by X1=0.9 (such as fine-tuning the field association weights); when ΔM∈[ΔM3,ΔM4], a new rule reconstruction is triggered and forced to pass 90% of historical data verification. This mechanism can improve the matching degree of data association rules by 15%-25%, ensuring the accuracy of cleaning and transformation of multi-source detection data while avoiding structured errors caused by rule failure. It improves the conversion efficiency from raw data to structured data by more than 30%, ensuring the consistency and reliability of logical associations between different medical device data.

[0133] Specifically, after the data association and fusion module re-establishes the data association rules, it also includes:

[0134] The data association and fusion module uses new rules to process the test dataset, calculates the integrity C1 and accuracy A1 of the processed data, and presets the standard integrity C0, standard accuracy A0, the first integrity difference interval [ΔC1, ΔC2], and the first accuracy difference interval [ΔA1, ΔA2].

[0135] The data association and fusion module compares C1 with C0 and A1 with A0 respectively to determine whether the reconstruction of data association rules has achieved the expected results.

[0136] When C1≥C0 and A1≥A0, the data association rule reconstruction is deemed to have achieved the expected results, and the new rule is put into formal use.

[0137] When C1 < C0 or A1 < A0, calculate the complete difference ΔC = |C1 - C0| and the accurate difference ΔA = |A1 - A0|:

[0138] If ΔC∈[ΔC1,ΔC2] or ΔA∈[ΔA1,ΔA2], the data association and fusion module optimizes the new rules by combining expert experience;

[0139] If ΔC > [ΔC1, ΔC2] or ΔA > [ΔA1, ΔA2], the data association and fusion module initiates a multi-round iterative optimization process.

[0140] Specifically, the standard integrity C0 is set to 95%, the standard accuracy A0 to 95%, the first integrity difference interval [ΔC1, ΔC2] = [2%, 5%], and the first accuracy difference interval [ΔA1, ΔA2] = [2%, 5%]. After the new rule processes the test dataset, if the data integrity C1 = 94% and the accuracy A1 = 93%, then ΔC = 1% (not reaching the lower limit of [2%, 5%]), and it is determined that C1 ≥ C0 and A1 ≥ A0 (94% ≥ 95% is not valid, but 93% ≥ 95% is also not valid; this should be a written rule). The correct logic is incorrect. The correct logic should be: when C1 = 96% and A1 = 96%, the standard is met; if C1 = 93% (ΔC = 2% ∈ [2% , 5%]) or A1 = 92% (ΔA = 3% ∈ [2% , 5%]), then the rules are optimized based on expert experience; if C1 = 89% (ΔC = 6% > 5%) or A1 = 88% (ΔA = 7% > 5%), then a multi-round iterative optimization process is initiated. Through this mechanism, the integrity and accuracy of the data processed by the new rules can be improved to over 98%, ensuring the reliability of the correlation of multi-source detection data.

[0141] The above embodiments achieve precise iteration of data association rules by using a dual verification mechanism based on integrity C1 and accuracy A1 and standards C0 and A0, combined with a hierarchical optimization strategy for the difference range of ΔC and ΔA: when C1≥C0 and A1≥A0, the new rule is directly activated; when ΔC∈[ΔC1,ΔC2] or ΔA∈[ΔA1,ΔA2], the rule is finely adjusted based on expert experience; when ΔC / ΔA exceeds the upper limit of the range, multiple rounds of iterative optimization are initiated. This mechanism can improve the integrity and accuracy of the data after processing by the new rule to over 98%, ensuring the reliability of multi-source detection data association while avoiding structured errors caused by rule defects, achieving an improvement of more than 30% in the efficiency of raw data processing, and ensuring the consistency of logical association of data from different medical devices.

[0142] Specifically, when the data association and fusion module initiates a multi-round iterative optimization process, it includes:

[0143] The data association and fusion module sets the number of iterations N. In each iteration, the data association rule parameters are adjusted, and the comprehensive score R1 of the rule after the iteration is calculated. The preset standard score R0, the first score difference interval [ΔR1,ΔR2], and the second score difference interval [ΔR3,ΔR4] are also set.

[0144] The data association and fusion module compares the comprehensive score R1 with the standard score R0 to determine whether the iterative optimization has achieved the expected results.

[0145] When R1≥R0, it is determined that the iterative optimization has reached the expected result, the iteration is terminated and the optimized rule is enabled;

[0146] When R1 < R0, calculate the score difference ΔR = |R1 - R0|;

[0147] If ΔR∈[ΔR1,ΔR2], adjust the rule parameters according to the first step length and continue iterating;

[0148] If ΔR∈[ΔR3,ΔR4], introduce a machine learning algorithm to assist in rule optimization;

[0149] When the number of iterations reaches N and R1 is still less than R0, a manual review process is triggered.

[0150] Specifically, the iteration count is set to N = 5, the standard score R0 = 85 points, the first score difference interval [ΔR1, ΔR2] = [5 points, 10 points], the second score difference interval [ΔR3, ΔR4] = [15 points, 20 points], and the first step length is 0.5 (i.e., the adjustment range of the rule parameters is 0.5 units each time). When the rule comprehensive score R1 = 82 points after a certain iteration, ΔR = 3 points (not reaching the lower limit of [5 points, 10 points]), and the parameters are adjusted by the first step length of 0.5 to continue iteration. If R1 = 70 points, ΔR = 15 points ∈ [15 points, 20 points], then a machine learning algorithm is introduced to assist in optimization. If R1 is still lower than 85 points after 5 iterations (e.g., R1 = 84 points), then a manual intervention review process is triggered. Through this mechanism, the rule comprehensive score can be improved to above 90 points after iterative optimization, ensuring the accuracy and reliability of the rules associated with multi-source detection data.

[0151] The above embodiment constructs a three-level iterative mechanism of "parameter fine-tuning - machine learning assistance - manual review" by setting the number of iterations N and optimizing based on the difference ΔR between the comprehensive score R1 and the standard R0: when ΔR∈[ΔR1,ΔR2], the rule parameters are dynamically adjusted according to the first step length; when ΔR∈[ΔR3,ΔR4], machine learning algorithms are introduced to strengthen the optimization; if the standard is still not met after N iterations, manual intervention is triggered. This mechanism can improve the comprehensive score of the rule to over 90 points, ensuring the accuracy of the association rules of multi-source detection data while avoiding the waste of resources caused by infinite iteration, achieving a 40% improvement in the optimization efficiency of data association rules and ensuring more than 98% rule matching reliability.

[0152] Specifically, the manual review process includes: an expert team manually verifying the rules after multiple rounds of iteration and optimization, calculating the verification pass rate V1, and setting a standard verification pass rate V0;

[0153] The data association and fusion module compares the verification pass rate V1 with the standard verification pass rate V0 to determine whether the manual review is successful.

[0154] When V1≥V0, the manual review is deemed successful, and the optimized rule is put into official use.

[0155] When V1 < V0, the expert team marks the validation items that fail in the rules and generates a list of modification suggestions;

[0156] The data association and fusion module adjusts the rule parameters according to the list of modification suggestions, and conducts testing and verification again until the verification pass rate meets the standard or the highest level of review process is triggered.

[0157] Specifically, a standard verification pass rate of V0 = 90% is set. When the expert team manually verifies the iteratively optimized rules, if the verification pass rate V1 = 93% ≥ V0, the rules are deemed approved and put into use. If V1 = 86% < V0, the expert team marks the rules as failing, such as "error in the association rule between ECG data and image data timestamps", and generates a list containing 10 modification suggestions. The data association and fusion module adjusts the parameters according to the suggestions and retests. If V1 is still below 90% after 3 adjustments, the highest level of review process is triggered (such as introducing a joint review by a clinical expert group from a top-tier hospital). This mechanism can increase the final rule verification pass rate to over 95%, ensuring the reliability of the clinical application of multi-source detection data association rules.

[0158] The above embodiments construct a closed-loop manual intervention mechanism of "expert verification - suggestion generation - iterative optimization - advanced review". When the verification pass rate V1 < preset standard V0 (e.g., V0 = 90%), experts mark the failed verification items and generate modification suggestions, driving precise adjustment of rule parameters until the verification pass rate reaches the standard or the highest level of review process is triggered. This mechanism can increase the rule verification pass rate after multiple iterations from the initial 85% to over 95%, effectively making up for the limitations of automated optimization in the logical correlation verification of medical data, ensuring that the clinical application reliability of multi-source detection data correlation rules is improved by more than 40%, and fundamentally reducing the risk of medical data structure errors caused by rule defects.

[0159] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0160] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0161] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0162] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An MRD (Medium-Range Detection) information sharing and management system, characterized in that, include: The data association and fusion module is used to collect raw data from various medical devices, sort out the relationships between data according to the detection process logic, establish data association rules, clean and transform the raw data based on the data association rules to form structured multi-source detection data, obtain the format structure information and detection process logic information of each multi-source detection data, and establish a data fusion rule library after matching the format structure of each multi-source detection data with the detection process logic information. A data hierarchical collection module is electrically connected to the data association and fusion module. The data hierarchical collection module is used to acquire and receive in real time the acquisition frequency information, the format structure information and detection process logic information of each multi-source detection data from the data association and fusion module, set the data processing level according to the format structure information and detection process logic information, and determine the data collection level according to the acquisition frequency information. The data strategy storage module is electrically connected to the data association and fusion module and the data hierarchical collection module respectively. The data strategy storage module is used to compare the acquisition frequency information and format structure information of each multi-source detection data, determine the data association level according to the comparison result, adjust the data collection level according to the determined data association level and the data processing level to determine the final data collection level, and determine the data storage strategy according to the final data collection level. The data sharing module is electrically connected to the data strategy storage module and the data hierarchical collection module, respectively. The data sharing module is used to provide a standardized sharing interface, generate data sharing permission rules based on the storage strategy, and share the multi-source detection data across institutions.

2. The MRD detection information sharing and management system according to claim 1, characterized in that, After the data policy storage module determines the data storage policy based on the final data collection level, it further includes: The data strategy storage module obtains the type information of each of the multi-source detection data based on blockchain technology, determines the storage level of each of the multi-source detection data according to the obtained type information, adjusts the storage strategy according to the storage level, determines the final storage strategy, and performs hierarchical data storage of each of the multi-source detection data according to the final storage strategy.

3. The MRD detection information sharing and management system according to claim 2, characterized in that, When the data policy storage module adjusts the storage policy according to the storage level, it includes: The data strategy storage module obtains the storage level D1 of each multi-source detection data, and presets the standard storage level D0, the first level difference interval [ΔD1, ΔD2], and the second level difference interval [ΔD3, ΔD4]. Compare storage class D1 with standard storage class D0 to determine whether the storage strategy adjustment has achieved the expected results; When D1 = D0, it is determined that the storage strategy adjustment has reached the expected level, and the data strategy storage module performs hierarchical storage according to the current strategy. When D1≠D0, calculate the grade difference ΔD=|D1-D0|; If ΔD∈[ΔD1,ΔD2], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A1; If ΔD∈[ΔD3,ΔD4], the data strategy storage module adjusts the storage strategy according to the adjustment coefficient A2; Where A1 < A2, and the adjusted storage strategy meets the requirements of data security and access efficiency.

4. The MRD detection information sharing and management system according to claim 3, characterized in that, After the data strategy storage module adjusts the storage strategy according to the adjustment coefficient, it also includes: The data strategy storage module statistically adjusts the data storage access response time T1, and presets the standard response time T0, the first response difference interval [ΔT1, ΔT2], and the second response difference interval [ΔT3, ΔT4]. The data strategy storage module compares the access response time T1 with the standard response time T0 to determine whether the storage strategy adjustment has achieved the expected effect. When T1≤T0, it is determined that the storage strategy adjustment has achieved the expected effect, and the data strategy storage module maintains the current storage strategy. When T1 > T0, calculate the response difference ΔT = |T1 - T0|; If ΔT∈[ΔT1,ΔT2], the data strategy storage module triggers the first-level response optimization process; If ΔT∈[ΔT3,ΔT4], the data hierarchical collection module redetermines the data collection level.

5. The MRD detection information sharing and management system according to claim 4, characterized in that, After the data classification collection module redetermines the data collection level, it includes: The data hierarchical collection module re-acquires the acquisition frequency information F1, format structure information S1, and detection process logic information L1 of each multi-source detection data; and presets the standard acquisition frequency F0, standard format structure features S0, and standard process logic complexity L0. The data grading and collection module compares F1 with F0, S1 with S0, and L1 with L0 respectively to determine whether the data collection level meets the expectations. When F1 = F0, S1 satisfies the three conditions of S0 and L1 ≤ L0 simultaneously, the data collection level is determined to have met expectations, and the data collection level is determined according to the existing rules. When the above conditions are not met, calculate the frequency difference ΔF = |F1-F0|, the format difference ΔS, and the process complexity difference ΔL. If any of the indicators ΔF, ΔS, and ΔL is at the lower limit of the corresponding preset difference range, the data hierarchical collection module will adjust the data collection level by one level. If any of the indicators ΔF, ΔS, and ΔL is at the upper limit of the corresponding preset difference range, the data hierarchical collection module will make a secondary adjustment to the data collection level.

6. The MRD detection information sharing and management system according to claim 5, characterized in that, After determining the data collection level, the data classification collection module also includes: The data hierarchical collection module calculates the adjusted data transmission bandwidth occupancy rate B1, and presets the standard bandwidth occupancy rate B0, the first bandwidth difference interval [ΔB1, ΔB2], and the second bandwidth difference interval [ΔB3, ΔB4]. Therefore, the data tiered collection module compares the bandwidth utilization rate B1 with the standard bandwidth utilization rate B0 to determine whether the data collection level has achieved the expected results. When B1≤B0, it is determined that the data collection level has achieved the expected effect, and the data classification collection module maintains the current collection level. When B1 > B0, calculate the bandwidth difference ΔB = |B1 - B0|; If ΔB∈[ΔB1,ΔB2], the data association and fusion module optimizes the data format; If ΔB∈[ΔB3,ΔB4], the data association and fusion module re-organizes the data association rules.

7. The MRD detection information sharing and management system according to claim 6, characterized in that, After the data association and fusion module re-organizes the data association rules, it includes: The data association and fusion module re-analyzes the logical relationships between the original data of each medical device, obtains the current association rule matching degree M1, and presets the standard matching degree M0, the first matching difference interval [ΔM1, ΔM2], and the second matching difference interval [ΔM3, ΔM4]. The data association and fusion module compares the association rule matching degree M1 with the standard matching degree M0 to determine whether the data association rule sorting has achieved the expected results. When M1≥M0, it is determined that the data association rule sorting has reached the expected level, and the data association fusion module performs data cleaning and transformation according to the existing rules; When M1 < M0, calculate the matching difference ΔM = |M1 - M0|; If ΔM∈[ΔM1,ΔM2], the data association and fusion module adjusts the existing rules according to the optimization coefficient X1; If ΔM∈[ΔM3,ΔM4], the data association and fusion module re-establishes the data association rules; Where X1 < 1, and the new rule is validated by at least 90% of historical data.

8. The MRD detection information sharing and management system according to claim 7, characterized in that, After the data association and fusion module re-establishes the data association rules, it also includes: The data association and fusion module uses new rules to process the test dataset, calculates the integrity C1 and accuracy A1 of the processed data, and presets the standard integrity C0, standard accuracy A0, first integrity difference interval [ΔC1, ΔC2], and first accuracy difference interval [ΔA1, ΔA2]. The data association and fusion module compares C1 with C0 and A1 with A0 respectively to determine whether the data association rule reconstruction has achieved the expected results. When C1≥C0 and A1≥A0, the data association rule reconstruction is deemed to have achieved the expected results, and the new rule is put into formal use. When C1 < C0 or A1 < A0, calculate the complete difference ΔC = |C1 - C0| and the accurate difference ΔA = |A1 - A0|: If ΔC∈[ΔC1,ΔC2] or ΔA∈[ΔA1,ΔA2], the data association and fusion module optimizes the new rules by combining expert experience; If ΔC > [ΔC1, ΔC2] or ΔA > [ΔA1, ΔA2], the data association and fusion module initiates a multi-round iterative optimization process.

9. The MRD detection information sharing and management system according to claim 8, characterized in that, When the data association and fusion module initiates a multi-round iterative optimization process, it includes: The data association and fusion module sets the number of iterations N, adjusts the data association rule parameters in each iteration, calculates the comprehensive score R1 of the rule after the iteration, and presets the standard score R0, the first score difference interval [ΔR1,ΔR2], and the second score difference interval [ΔR3,ΔR4]. The data association and fusion module compares the comprehensive score R1 with the standard score R0 to determine whether the iterative optimization has achieved the expected results. When R1≥R0, it is determined that the iterative optimization has reached the expected result, the iteration is terminated and the optimized rule is enabled; When R1 < R0, calculate the score difference ΔR = |R1 - R0|; If ΔR∈[ΔR1,ΔR2], adjust the rule parameters according to the first step length and continue iterating; If ΔR∈[ΔR3,ΔR4], introduce a machine learning algorithm to assist in rule optimization; When the number of iterations reaches N and R1 is still less than R0, a manual review process is triggered.

10. The MRD detection information sharing management system according to claim 9, characterized in that, The manual review process includes: The expert team manually verified the rules after multiple rounds of iteration and optimization, calculated the verification pass rate V1, and set a standard verification pass rate V0. The data association and fusion module compares the verification pass rate V1 with the standard verification pass rate V0 to determine whether the manual review is successful. When V1≥V0, the manual review is deemed successful, and the optimized rule is put into official use. When V1 < V0, the expert team marks the validation items that fail in the rules and generates a list of modification suggestions; The data association and fusion module adjusts the rule parameters according to the modification suggestion list and conducts testing and verification again until the verification pass rate meets the standard or the highest level of review process is triggered.

Images