An intelligent hierarchical classification and privacy protection system for medical data

Abstract

The present invention belongs to the technical field of data security. A smart hierarchical classification and privacy protection system for medical data is provided. The system includes: a data receiving unit for receiving each piece of medical data that needs privacy protection, where the medical data contains corresponding disease type information and treatment method information; a data hierarchical classification unit for evaluating the potential application intensity level of the corresponding medical data based on the disease type information and treatment method, and classifying or grading the medical data according to the potential application intensity level; a privacy processing unit for determining a privacy processing plan for the medical data according to the classification or grading, and performing privacy processing on the medical data according to the privacy processing plan. The above solution of the present invention can predict the potential application intensity of each piece of medical data, and then adopt different privacy processing plans, so as to achieve a relative balance between the privacy protection and reasonable application of medical data.

Description

Technical Field

[0001] The present invention relates to the technical field of data security, and in particular, to an intelligent classification and privacy protection system for medical data. Background Art

[0002] In the digital age, the medical field is undergoing profound changes, and the value of medical data is becoming increasingly prominent.

[0003] On the one hand, medical data covers patients' medical histories, diagnosis results, treatment plans, genetic information, etc. These data play a huge role in disease research, clinical decision support, medical quality assessment, and drug development. For example, by analyzing a large amount of medical data of patients, researchers can discover the potential laws of diseases and develop more effective treatment methods; medical institutions can also optimize the medical service process based on these data to improve the diagnosis and treatment efficiency. On the other hand, the high sensitivity and privacy of medical data cannot be ignored. Once these data are leaked, it will have serious negative impacts on patients, such as personal privacy exposure, social discrimination, economic losses, etc.

[0004] Therefore, how to achieve a relative balance between the privacy protection and reasonable application of medical data is a technical problem that urgently needs to be solved at present. Summary of the Invention

[0005] In view of this, the present invention provides an intelligent classification and privacy protection system for medical data, an electronic device, a computer storage medium, and a computer program product to solve the above technical problems.

[0006] The present invention provides an intelligent classification and privacy protection system for medical data, which includes a data receiving unit, a data classification unit, and a privacy processing unit; the data receiving unit is used to receive each medical data that needs privacy protection, and the medical data contains corresponding disease type information and treatment method information; the data classification unit is used to evaluate the potential application intensity level of the corresponding medical data according to the disease type information and the treatment method information, and classify or grade the medical data according to the potential application intensity level; the privacy processing unit is used to determine the privacy processing scheme for the medical data according to the classification or grading, and perform privacy processing on the medical data according to the privacy processing scheme.

[0007] Optionally, the receiving of each medical data that needs privacy protection includes: after the data receiving unit receives the medical data, it parses the medical data and determines whether the parsing result contains identity information. If the parsing result contains identity information, it is determined that the medical data needs privacy protection.

[0008] Optionally, the data classification and grading unit includes an associated information query subunit, a feature extraction subunit, and an application intensity analysis subunit; the associated information query subunit searches in a number of specified data sources according to the disease type information and the medical treatment means, and respectively obtains a first search result corresponding to the disease type information and a second search result corresponding to the disease type information and the medical treatment means information; the feature extraction subunit extracts features from the first search result and the second search result to obtain disease features; the application intensity analysis subunit calls a deep analysis model to process the disease features to obtain the potential application intensity level of the evaluated medical data.

[0009] Optionally, the extracting features from the first search result and the second search result to obtain disease features includes: the feature extraction subunit uses a feature extraction model to extract from the first search result the first result quantity corresponding to the disease type information and the result time span; uses a semantic analysis component to extract from the first search result multiple disease onset principles corresponding to the disease type information, counts the second result quantity of the disease onset principles whose result quantity is higher than a preset value, and extracts from the second search result the medical treatment means data set corresponding to the disease type information, and analyzes the evolution trend of the medical treatment means according to the medical treatment means data set; vectorizes the first result quantity, the result time span, the second result quantity, and the evolution trend of the medical treatment means to obtain the disease features.

[0010] Optionally, the application intensity analysis subunit calling a deep analysis model to process the disease features to obtain the potential application intensity level of the evaluated medical data includes: the application intensity analysis subunit calls a deep analysis model to process the disease features to obtain the initial potential application intensity level of the evaluated medical data; the application intensity analysis subunit also obtains the medical data sharing history of the ownership institution of the medical data, statistically obtains the sharing scale of data that is the same as or similar to the medical data according to the medical data sharing history, and determines an optimization coefficient according to the sharing scale; uses the optimization coefficient to optimize the initial potential application intensity level to the potential application intensity level.

[0011] Optionally, the classifying or grading the medical data according to the potential application intensity level includes: matching the corresponding grading label / classification label according to the potential application intensity level, and associating the grading label / classification label with the medical data.

[0012] Optionally, determining a privacy processing solution for the medical data according to the grading or the classification includes: determining the privacy processing solutions for the medical data according to the levels of the potential application intensity corresponding to the grading labels or the classification labels, and different privacy processing solutions are implemented by erasing the identity information in the medical data to different extents and / or encrypting the medical data by using encryption algorithms with different encryption levels.

[0013] The present invention also discloses an electronic device, which is applied to the system as described in any one of the preceding items, and includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor.

[0014] The present invention also discloses a computer storage medium, which is applied to the system as described in any one of the preceding items, and the computer-readable storage medium stores a computer program.

[0015] The present invention also discloses a computer program product, which is applied to the system as described in any one of the preceding items, and the computer program contains a number of computer codes.

[0016] The above solution of the present invention can predict the potential application intensity of each piece of medical data, and then adopt different privacy processing solutions, so as to achieve a relative balance between the privacy protection and the reasonable application of medical data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 is a schematic structural diagram of an intelligent grading and classification and privacy protection system for medical data disclosed in an embodiment of the present invention.

[0019] Figure 2 is a schematic structural diagram of a data grading and classification unit disclosed in an embodiment of the present invention.

[0020] Figure 3 is a schematic structural diagram of a feature extraction sub-unit disclosed in an embodiment of the present invention.

[0021] Figure 4 is a schematic structural diagram of an application intensity analysis sub-unit disclosed in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The following specific embodiments illustrate the implementation manners of the present application. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0023] In addition, the technical features involved in different implementation manners of the present application described below can be combined with each other as long as they do not conflict with each other.

[0024] As Figure 1 shown, an embodiment of the present invention discloses an intelligent grading and classification and privacy protection system for medical data, and the system includes a data receiving unit, a data grading and classification unit, and a privacy processing unit.

[0025] The data receiving unit is used to receive each medical data that needs privacy protection, and the medical data contains corresponding disease type information and treatment method information.

[0026] The data receiving unit is the key interface for the entire system to interact with external data. It is used to accurately connect to various medical data sources, and then receive multiple medical data. The medical data comes from, for example, the patient diagnosis and treatment information in the hospital's electronic medical record system, the test reports of medical testing institutions, and also includes the monitoring data from remote medical devices, wearable health monitoring devices, etc. In addition to the patient's identity information (name, gender, age, contact information, native place, family address, past medical history, etc.), these medical data also contain disease type information and treatment method information. The disease type information is, for example, pneumonia, thyroid cancer, influenza, etc., and the treatment method information includes taking a certain drug (capsule, tablet), surgery, liquid nitrogen freezing, laser, chemotherapy, targeted drug, etc.

[0027] The data grading and classification unit is used to evaluate the corresponding potential application intensity level of the medical data according to the disease type information and the treatment method information, and classify or grade the medical data according to the potential application intensity level.

[0028] Since different medical data play different roles in disease research, clinical decision-making, medical quality assessment, etc., there are obvious differences in their potential application intensities. Therefore, the data grading and classification unit of the present invention uses a suitable evaluation method to evaluate the potential application intensity level of medical data based on disease type information and diagnosis and treatment means information. The potential application intensity level in the present invention refers to the intensity of cross-system and cross-region application of medical data. For example, if the probability of a certain medical data being applied across more systems (or institutions) and more regions (countries, provinces, cities, etc.) is relatively high, its corresponding potential application intensity level is higher, and vice versa. The systems mentioned here refer to different medical data management systems, such as between different medical institutions, between the official medical data management system and the third-party commercial medical data application system, etc.

[0029] For example, for the research data of rare diseases such as amyotrophic lateral sclerosis (ALS), since it is crucial for revealing the pathogenesis of the disease and exploring effective treatment methods, the potential application intensity level is high, that is, the greater the value of this medical data, the greater the possibility of being applied by more institutions in various ways. For common colds, the treatment methods are already very standardized and perfect. Generally speaking, the potential application intensity level of such medical data (such as symptom records, simple blood routine test results, etc.) is relatively low, that is, the smaller the value of this medical data, the smaller the possibility of being applied by more institutions in various ways.

[0030] According to the potential application intensity level obtained by evaluation, the medical data is reasonably graded or classified. For example, the data with high application intensity is classified as level one and used for in-depth scientific research and precise diagnosis and treatment of difficult diseases; the data with low application intensity is classified as level three and mainly used for daily medical statistical analysis.

[0031] The present invention abandons the traditional extensive data processing mode, enabling data with different values and sensitivities to be sorted out in an orderly manner, and can significantly improve the pertinence and efficiency of subsequent data processing and use.

[0032] The privacy processing unit is used to determine the privacy processing scheme for the medical data according to the grading or classification, and perform privacy processing on the medical data according to the privacy processing scheme.

[0033] Different classifications / gradings correspond to different potential application intensity levels, and different privacy processing schemes are adopted accordingly. The privacy processing schemes include, but are not limited to, different degrees of erasure of identity information, encryption methods for medical data (different encryption methods correspond to different encryption intensities), etc.

[0034] The above solution of the present invention can predict the potential application intensity of each piece of medical data, and then adopt different privacy processing solutions, so as to achieve a relative balance between the privacy protection and reasonable application of medical data.

[0035] Optionally, the receiving of each piece of medical data that needs privacy protection includes: after the data receiving unit receives the medical data, it parses the data and determines whether the parsing result contains identity information. If it contains identity information, it determines that the medical data is the medical data that needs privacy protection.

[0036] In this embodiment, the medical data received by the data receiving unit comes from a wide range of sources, covering hospital information systems, medical testing institution reports, remote medical devices, and wearable health monitoring devices, etc. Some medical data may not contain the identity information of patients, so there is no need for privacy processing. For this reason, the present invention is configured such that when the data receiving unit receives medical data, it will immediately parse the data (including operations such as data format identification and data content extraction), and determine whether it contains identity information. If it contains identity information, it determines that the medical data is data that needs privacy protection, and performs privacy processing on it before subsequent use.

[0037] By the above method, the number of medical data that needs privacy protection can be effectively reduced, thereby reducing the data processing load.

[0038] Optionally, as Figure 2 shown, the data classification and grading unit includes an associated information query subunit, a feature extraction subunit, and an application intensity analysis subunit; the associated information query subunit searches in several specified data sources according to the disease type information and the treatment means, and respectively obtains a first search result corresponding to the disease type information and a second search result corresponding to the disease type information and the treatment means information; the feature extraction subunit extracts features from the first search result and the second search result to obtain disease features; the application intensity analysis subunit calls a deep analysis model to process the disease features to obtain the potential application intensity level of the evaluated medical data.

[0039] In this embodiment, as Figure 2 shown, the data classification and grading unit further includes the above three subunits. Among them: the associated information query subunit searches in several specified data sources such as an authoritative medical database, a professional medical research literature database, and a case database of large medical institutions according to the received disease type information and treatment means information. For example, when the disease type information is "leukemia" and the treatment means is "bone marrow transplantation", the associated information query subunit searches in the above data sources with "leukemia" and "leukemia" + "bone marrow transplantation" respectively.

[0040] The feature extraction subunit extracts key features, i.e., the above-mentioned disease features, from the above search results. The application intensity analysis subunit then calls the in-depth analysis model to process the extracted disease features, thereby obtaining the potential application intensity level of the medical data. The in-depth analysis model can be a model constructed based on machine learning algorithms, such as neural network models, general large models (DeepSeek, GPT, BERT, etc.).

[0041] Optionally, the feature extraction of the first search result and the second search result to obtain disease features includes: the feature extraction subunit uses a feature extraction model to obtain from the first search result the first result quantity corresponding to the disease type information and the result time span; uses a semantic analysis component to obtain from the first search result multiple pathogenesis principles corresponding to the disease type information, counts the second result quantity of the pathogenesis principles with the result quantity higher than the preset value, and obtains from the second search result a dataset of diagnosis and treatment means corresponding to the disease type information, and analyzes the evolution trend of the diagnosis and treatment means according to the dataset of diagnosis and treatment means; vectorizes the first result quantity, the result time span, the second result quantity, and the evolution trend of the diagnosis and treatment means to obtain the disease features.

[0042] In this embodiment, as Figure 3 shown, the feature extraction subunit uses a feature extraction model (such as a convolutional network) to count the quantity of relevant information, i.e., the first result quantity and the result time span, from the first search result corresponding to the disease type information. The two are used to characterize the degree to which the disease is studied and discussed. Generally speaking, the more the first result quantity and the larger the result time span, the higher the degree to which the disease is studied and discussed. Correspondingly, it also indicates a higher probability that the disease belongs to a common disease and the treatment method is mature; otherwise, it is lower.

[0043] A semantic analysis component is also embedded in the feature extraction subunit. This component can perform in-depth semantic analysis on the first search result to extract multiple pathogenesis principles corresponding to the disease type information. For example, for diabetes, the pathogenesis principles may include insulin resistance, pancreatic islet β cell dysfunction, etc. Determine the pathogenesis principles with the result quantity higher than the preset value (such as 500) as the pathogenesis principles with higher confidence, and count the second result quantity of all pathogenesis principles with higher confidence.

[0044] The semantic analysis component also extracts a dataset of diagnosis and treatment methods corresponding to the disease type information from the second search result, and analyzes the evolution trend of the diagnosis and treatment methods based on the dataset of diagnosis and treatment methods. For example, if the evolution trend of the diagnosis and treatment methods shows that the quantity curves of various diagnosis and treatment methods tend to be stable (for example, the quantity curves are in a generally horizontal state), it indicates that the treatment methods for this disease have been very mature. When the quantity curves of various diagnosis and treatment methods do not tend to be stable or are even in a chaotic state of interlacing, it indicates that the treatment methods for this disease are still in the research stage (for example, this disease is a certain new disease, or a difficult and complicated disease that has never been conquered).

[0045] Vectorize the above-mentioned first result quantity, result time span, second result quantity, and evolution trend of diagnosis and treatment methods obtained by extraction and analysis, that is, obtain disease characteristics in the form of a numerical vector.

[0046] Optionally, as Figure 4 shown, the application intensity analysis subunit calls a depth analysis model to process the disease characteristics, and obtains the potential application intensity level of the medical data to be evaluated, including: the application intensity analysis subunit calls a depth analysis model to process the disease characteristics, and obtains the initial potential application intensity level of the medical data to be evaluated; the application intensity analysis subunit also obtains the medical data sharing history of the ownership institution of the medical data, statistically obtains the sharing scale of data that is the same as or similar to the medical data according to the medical data sharing history, and determines an optimization coefficient according to the sharing scale; uses the optimization coefficient to optimize the initial potential application intensity level to the potential application intensity level.

[0047] In this embodiment, the well-trained depth analysis model performs prediction processing on the previously extracted and vectorized disease characteristics, and preliminarily evaluates the initial potential application intensity level of the medical data of this disease. For example, for a rare disease, if the number of relevant research literatures is small (the quantity of the first result is small) but the time span is long (indicating that it has been concerned for a long time but the progress is slow), the pathogenesis is complex (the quantity of the second result is large), and the diagnosis and treatment methods are not yet stable (the evolution trend of the diagnosis and treatment methods shows that the quantities of various diagnosis and treatment methods have not tended to be stable), then the initial potential application intensity level of the medical data of this disease is preliminarily evaluated as high, that is, relevant institutions have a relatively large application demand for this medical data, because there is still a large research space for the pathogenesis and diagnosis and treatment methods of the disease corresponding to this medical data, and the research value and potential influence of this medical data are relatively large.

[0048] The above-mentioned initial potential application intensity level is analyzed without considering the willingness of the medical data ownership institution to share medical data. In fact, the willingness of the medical data ownership institution to share medical data also has a direct and significant impact on the potential application intensity level of medical data. In this regard, the present invention is further configured to obtain the medical data sharing history of the medical data ownership institution, and statistically obtain the sharing scale of data that is the same as or similar to the medical data according to the medical data sharing history. Among them, the same or similar data refers to the data belonging to the above-mentioned initial potential application intensity level, that is, analyzing how much medical data of this initial potential application intensity level the ownership institution has shared; the sharing scale is mainly determined by factors such as the total amount of shared data, the number of sharing batches, and the number of sharing objects. Then, an optimization coefficient is determined according to the sharing scale (for example, through multiplication operation), and this optimization coefficient is actually used to reflect the degree to which the ownership institution hinders the sharing of medical data. Obviously, the larger the sharing scale, the lower the above-mentioned hindrance degree, and the closer the corresponding optimization coefficient is to 1, that is, the less the initial potential application intensity level is lowered; while the smaller the sharing scale, the higher the above-mentioned hindrance degree, and the corresponding optimization coefficient is less than 1, that is, the more the initial potential application intensity level is increased.

[0049] Therefore, through the optimization coefficient, the initially obtained potential application intensity level can be adjusted to be closer to the actual situation, which is beneficial to subsequent more appropriate grading or classification of medical data.

[0050] Optionally, the grading or classifying the medical data according to the potential application intensity level includes: matching the corresponding grading label / classification label according to the potential application intensity level, and associating the grading label / the classification label with the medical data.

[0051] In this embodiment, after determining the corresponding potential application intensity level, the corresponding grading label or classification label can be matched according to the preset corresponding relationship and associated with the medical data. Subsequently, when in use, different privacy processing schemes can be determined according to this label.

[0052] Optionally, the determining the privacy processing scheme for the medical data according to the grading or the classification includes: determining the privacy processing scheme for each medical data according to the level of the potential application intensity level corresponding to the grading label or the classification label, and different privacy processing schemes are realized by erasing the identity information in the medical data to different degrees and / or encrypting the medical data using encryption algorithms with different encryption levels.

[0053] In this embodiment, a personalized privacy processing solution for the medical data can be determined according to the level of potential application intensity. For example, when the potential application intensity level is relatively high, the probability that it is used by more institutions and across more regions is higher. At this time, more content in its identity information can be erased, such as erasing all of the name, gender, age, contact information, place of origin, home address, past medical history, etc. When the potential application intensity level is relatively low, only the name, contact information, place of origin, home address, past medical history, etc. are erased.

[0054] The privacy processing of medical data can also be reflected by different encryption algorithms, which can reduce the probability of medical data being leaked during the process of transmission and use. For example, when the potential application intensity level is relatively high, encryption algorithms with better encryption performance such as AES and RSA are adopted. When the potential application intensity level is relatively low, encryption algorithms with slightly worse encryption performance such as EDS and 3DES are adopted.

[0055] The above is only for illustrative purposes and is not used to limit the protection scope of the present invention.

[0056] An embodiment of the present invention also discloses an electronic device, which is applied to the system as described in any one of the previous items, and includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor.

[0057] An embodiment of the present invention also discloses a computer storage medium, which is applied to the system as described in any one of the previous items, and the computer-readable storage medium stores a computer program.

[0058] An embodiment of the present invention also discloses a computer program product, which is applied to the system as described in any one of the previous items, and the computer program contains a number of computer codes.

[0059] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0060] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0061] The above-described embodiments merely represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. An intelligent classification and privacy protection system for medical data, characterized by: The system includes a data receiving unit, a data grading and classification unit, and a privacy processing unit; the data receiving unit is used to receive various medical data that need to be privacy protected, wherein the medical data contains corresponding disease type information and diagnosis and treatment means information; the data grading and classification unit is used to evaluate the corresponding potential application intensity level of the medical data according to the disease type information and the diagnosis and treatment means information, and to grade or classify the medical data according to the potential application intensity level; the privacy processing unit is used to determine a privacy processing scheme for the medical data according to the grading or classification, and to perform privacy processing on the medical data according to the privacy processing scheme; the potential application intensity level refers to the intensity of medical data being applied across systems and regions; The data classification unit includes a correlation information query subunit, a feature extraction subunit, and an application intensity analysis subunit; The associated information query subunit searches in several designated data sources according to the disease type information and the diagnosis and treatment means, and obtains a first search result corresponding to the disease type information and a second search result corresponding to the disease type information and the diagnosis and treatment means information respectively; the feature extraction subunit extracts features from the first search result and the second search result to obtain disease features; The application intensity analysis subunit calls the deep analysis model to process the disease characteristics to obtain the evaluated potential application intensity level of the medical data; The feature extraction of the first search result and the second search result to obtain disease features includes: the feature extraction subunit uses a feature extraction model to extract the first result quantity and result time span corresponding to the disease type information from the first search result; uses a semantic analysis component to extract multiple pathogenesis principles corresponding to the disease type information from the first search result, counts the number of second results of pathogenesis principles whose result number is higher than a preset value, and extracts a diagnosis and treatment means data set corresponding to the disease type information from the second search result, and obtains the evolution trend of diagnosis and treatment means according to the analysis of the diagnosis and treatment means data set; The first number of results, the time span of the results, the second number of results, and the evolution trend of the diagnosis and treatment methods are vectorized to obtain the disease characteristics.

2. The intelligent classification and privacy protection system for medical data according to claim 1, characterized in that: The receiving of each medical data requiring privacy protection includes: after the data receiving unit receives the medical data, the data receiving unit analyzes the medical data to determine whether the analysis result contains identity information; if the identity information is contained, the data is determined to be the medical data requiring privacy protection.

3. The intelligent classification and privacy protection system for medical data according to claim 2, characterized in that: The application strength analysis subunit calls the deep analysis model to process the disease characteristics to obtain the evaluated potential application strength level of the medical data, including: the application strength analysis subunit calls the deep analysis model to process the disease characteristics to obtain the evaluated initial potential application strength level of the medical data; the application strength analysis subunit also obtains the medical data sharing history of the institution that owns the medical data, obtains the sharing scale of data that is the same or similar to the medical data based on the medical data sharing history, and determines an optimization coefficient based on the sharing scale; and uses the optimization coefficient to optimize the initial potential application strength level to the potential application strength level.

4. The intelligent classification and privacy protection system for medical data according to claim 1, characterized in that: The grading or classifying the medical data according to the potential application strength level includes: obtaining a corresponding grading label / classification label according to the potential application strength level matching, and associating the grading label / classification label with the medical data.

5. The intelligent classification and privacy protection system for medical data according to claim 1, characterized in that: The determining of the privacy processing scheme for the medical data according to the grading or the classification includes: determining the privacy processing scheme for each medical data according to the potential application strength level corresponding to the grading label or the classification label, and different privacy processing schemes are achieved by erasing the identity information in the medical data to different degrees and / or encrypting the medical data using encryption algorithms with different encryption levels.

6. An electronic device, characterized in that: The system applied to any one of claims 1 to 5 comprises: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor.

7. A computer storage medium, characterized in that: Applied to the system according to any one of claims 1 to 5, the computer storage medium stores a computer program.

8. A computer program product, applied to the system according to any one of claims 1 to 5, characterized in that: The computer program includes several computer codes.

Images