Medical care health sensor data security protection method, system, medium and server

Abstract

The application provides a medical care health sensor data security protection method, system, medium and server, relates to the technical field of data security, and the method comprises the following steps: firstly, collecting historical data of a medical care health sensor and obtaining user information; then, extracting user information features and encoding; then, constructing a first matrix by using the encoding values of any feature, and forming a second matrix by using a block splicing algorithm; subsequently, segmenting the second matrix by using a quadtree segmentation algorithm, judging the data attribution of the segmented blocks, taking the segmented blocks as quadtree leaf nodes if the segmented blocks belong to the same first matrix, otherwise, updating the segmented blocks into the second matrix and continuing the segmentation; finally, performing path encoding on the leaf nodes, performing scrambling processing on the encoding results as labels, and obtaining and storing the final processing results. The application can efficiently and safely process data.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of data security, in particular to a medical and health sensor data security protection method, system, medium and server. BACKGROUND

[0002] With the rapid development of Internet of Things (IoT) and medical informatization technology, medical and health sensors (such as wearable devices, home medical monitoring instruments, etc.) have been widely used in remote medical treatment, health management, chronic disease monitoring and other fields. These sensors continuously collect physiological data (such as heart rate, blood pressure, blood glucose), behavior data (such as movement trajectory, sleep pattern) and environmental data (such as temperature and humidity, air quality) of users, providing a basis for personalized medical treatment and health management.

[0003] A full-process data security protection system is provided in Chinese patent application publication No. CN115664830A. When a server loads and executes the computer program corresponding to the patent, the server implements the following steps: obtaining user data uploaded by a target user, encrypting the user data according to an encryption algorithm to obtain encrypted data, when a third party uses the encrypted data, sampling the encrypted data according to a preset rule to obtain sampled data, encrypting the sampled data according to the security level encryption strategy corresponding to each field based on the security level encryption strategy corresponding to different fields in the encrypted data, and providing the encrypted sampled data to the third party for display.

[0004] However, encryption algorithms usually need to perform a series of mathematical operations on data, such as permutation, substitution, XOR operation, etc., to convert the original data into ciphertext. These operations consume certain computing resources and time, especially when dealing with large-scale data, the encryption process will become more time-consuming. For example, for a data set containing a large number of patient health records, encrypting each data item will significantly increase the time cost of data processing. SUMMARY

[0005] In order to be able to process data efficiently and safely, the application provides a medical and health sensor data security protection method, system, medium and server.

[0006] In a first aspect, the application provides a medical and health sensor data security protection method, which adopts the following technical solution:

[0007] The medical and health sensor data security protection method comprises:

[0008] Historical data collection: collecting historical data of medical and health sensors, and obtaining user information of the medical and health sensors;

[0009] Constructing data matrix: extract multiple features of user information, encode each feature separately; construct a first matrix for each medical and health sensor, the diagonal elements of the first matrix are the encoding values of any one feature, the non-diagonal elements of the first matrix are the historical data from the same medical and health sensor, and when the historical data is insufficient to fill the non-diagonal elements of the first matrix, a specific character is used to fill the non-diagonal elements of the first matrix, and the same label is added to the historical data belonging to the same first matrix; using block splicing algorithm, splicing the first matrix into a second matrix;

[0010] Data matrix segmentation: using quadtree segmentation algorithm to segment the second matrix to obtain multiple segmentation blocks;

[0011] Attribute judgment: judging whether the historical data in the current segmentation block belongs to the same first matrix according to the label, if yes, executing the step of leaf node adding; if not, executing the step of data matrix updating;

[0012] Leaf node adding: taking the historical data in the current segmentation block as a leaf node in the quadtree structure;

[0013] Data matrix updating: updating the current segmentation block to the second matrix, and executing the step of data matrix segmentation;

[0014] Encoding: path encoding processing is performed on the leaf node to obtain an encoding result;

[0015] Scrambling: taking the encoding result as the label of the leaf node, and performing scrambling processing on the leaf node to obtain and store the processing result.

[0016] The application firstly extracts a plurality of features in user information, encodes the features to obtain an encoded value, then constructs a first matrix by using the encoded value, and constructs a second matrix by using a block splicing algorithm. The application not only retains the original features of the data, but also increases the complexity by introducing diagonal elements. The application improves the security of the data through such disorder processing. Subsequently, the application uses a quadtree segmentation algorithm to segment the second matrix, so that each segmentation block contains a meaningful data subset. Subsequently, the application realizes dynamic recognition of data attribution by judging whether the historical data in the current segmentation block belongs to the same first matrix. When the data in the segmentation block belongs to the same first matrix, it is directly added as a leaf node; otherwise, the segmentation block is updated to a second matrix and continues to be segmented to realize the layer-by-layer refinement of the data, and improve the efficiency and accuracy of data processing. Subsequently, the application performs path encoding processing on the leaf node, converts the complex data structure into an encoded result easy to manage and transmit, thereby simplifying the data processing process and increasing the security of the data. Finally, the application takes the encoded result as the label of the leaf node and performs shuffling processing on the leaf node, thereby breaking the original association between the data, making it difficult for attackers to infer the original data from the processing result, and significantly improving the security of the data.

[0017] Optionally, the step of constructing the data matrix further comprises:

[0018] Number judgment: judging whether the number of the features is less than the number of the medical care health sensors, if yes, performing the step of increasing the features; if no, performing the step of constructing the data matrix;

[0019] Increasing the features: statistically analyzing the behavior data of the user to obtain new features, merging the new features with the features in the step of feature extraction, and calculating the correlation between any two features by using a cosine similarity algorithm, denoted as a first correlation.

[0020] In the step of constructing the data matrix, the first matrix is spliced in the order of the first correlation from large to small.

[0021] The application determines whether the number of features is less than the number of medical and health sensors, so as to determine whether the number of features is sufficient before matrix construction. When the number of features is less than the number of medical and health sensors, the step of adding features is performed, so that each medical and health sensor can have sufficient features for association and analysis. Subsequently, the application obtains new features by statistically analyzing the behavior data of the user, which can reflect the user's usage habits, behavior patterns and the like, thereby providing additional dimensions for data security protection. Subsequently, the application combines the new features with the original features to enrich the feature set. Subsequently, the application calculates the correlation between any two features using a cosine similarity algorithm, denoted as a first correlation, to reveal the inherent relationship between the features. The application splices the first matrix in the order of the first correlation from large to small, so that the features with high correlation are adjacent or close to each other in the second matrix, and the four-tree structure has high correlation between adjacent nodes. That is, in the process of four-tree construction, high-correlation features arranged adjacently or close to each other are preferentially divided into the same node or adjacent nodes. This spatial topology structure can naturally form a feature clustering effect when the four-tree is split, that is, a feature subset with similar semantics or statistical correlation is automatically merged into the same sub-tree. When performing a region query or a nearest neighbor search, the search path can be predicted based on the correlation strength between nodes, thereby improving the retrieval efficiency.

[0022] Optionally, after the step of adding features, the method further comprises:

[0023] Historical data and feature correlation analysis: the cosine similarity algorithm is used to calculate the correlation between the historical data of each medical and health sensor and the i-th feature, denoted as a second correlation.

[0024] In the step of constructing the data matrix, the historical data of each medical and health sensor is filled into the corresponding first matrix in the order of the second correlation from large to small.

[0025] The application uses the cosine similarity algorithm to calculate the correlation between the historical data of each medical and health sensor and the i-th feature, denoted as a second correlation, which can quantify the degree of association between the historical data of each medical and health sensor and the feature. In the step of constructing the data matrix, the historical data of each medical and health sensor is filled into the corresponding first matrix in the order of the second correlation from large to small, so that the first matrix not only retains the original information of the historical data, but also highlights the historical data closely related to the features in the user information.

[0026] Optionally, the step of constructing the data matrix further comprises:

[0027] Data analysis: set the normal range of the output data of each medical and health sensor;

[0028] Range judgment: judging whether the jth non-diagonal element in each first matrix falls within the normal range in turn, if yes, retaining the jth non-diagonal element in the first matrix; if no, replacing the jth non-diagonal element with a specific character;

[0029] Iteration: taking the j+1th non-diagonal element as a new jth non-diagonal element, and performing the step of range judgment until all non-diagonal elements in the first matrix are traversed.

[0030] The application first sets the normal range of each medical and health sensor output data, then checks whether the jth non-diagonal element in the first matrix falls within the normal range, if the non-diagonal element is not within the normal range, the jth non-diagonal element is updated to a specific character. By filling the specific character, the application can replace the abnormal data points, maintain the accuracy and reliability of the data in the first matrix; otherwise, the jth non-diagonal element in the first matrix is retained, and the j+1th non-diagonal element is taken as a new jth non-diagonal element, and then the step of range judgment is continued. By using the above technical solution, the application can filter out the non-standard data as much as possible, and improve the accuracy and reliability of data processing.

[0031] Optionally, after performing the step of attribute judgment, before performing the step of data matrix updating, the application further comprises:

[0032] Segmentation block judgment: judging whether the number of historical data contained in the current segmentation block is less than a preset minimum number threshold, if yes, performing the step of data matrix updating; if no, performing the step of segmentation block updating;

[0033] Segmentation block updating: calculating the covariance matrix of the non-diagonal elements in the current segmentation block, and calculating the maximum eigenvalue of the covariance matrix and the corresponding eigenvector, segmenting the current segmentation block according to the direction of the eigenvector to obtain a first segmentation sub-block and a second segmentation sub-block, sequentially updating the first segmentation sub-block and the second segmentation sub-block to the current segmentation block, and performing the step of data matrix updating.

[0034] The application first checks whether the number of historical data contained in the current segmentation block is less than a preset minimum number threshold. If yes, the step of data matrix updating is performed; otherwise, the covariance matrix of the non-diagonal elements in the current segmentation block is calculated, and the maximum eigenvalue and the eigenvector corresponding to the maximum eigenvalue are extracted. The covariance matrix reflects the correlation between the data, and the eigenvalue and the eigenvector reveal the main change direction and amplitude of the data. Subsequently, the current segmentation block is segmented according to the direction of the eigenvector to obtain a first segmentation sub-block and a second segmentation sub-block, which can cut the data block in the main change direction, helping to more accurately cut the current segmentation block. Subsequently, the first segmentation sub-block and the second segmentation sub-block are updated to the current segmentation block in turn, and the step of data matrix updating is performed. Through covariance analysis and eigenvector segmentation, the application can more accurately identify the key information and patterns in the data, segment the current segmentation block in the direction with the greatest difference, and thus improve the efficiency of segmentation.

[0035] Optionally, after the step of data matrix updating, the method further comprises:

[0036] Adding empty nodes: obtaining a quadtree structure, adding a plurality of empty nodes to the kth layer leaf nodes in the quadtree structure, updating the added empty nodes to the kth layer leaf nodes in the quadtree structure to obtain an expanded quadtree structure;

[0037] In the encoding step, the path encoding processing is performed on each layer leaf node in the expanded quadtree structure to obtain a new encoding result.

[0038] The application first obtains the current quadtree structure and adds a plurality of empty nodes to the kth layer leaf nodes. These empty nodes are then updated to the kth layer leaf nodes to form an expanded quadtree structure. By adding empty nodes and updating them to new leaf nodes, the coding space of the quadtree structure can be effectively increased without changing the original data structure. Subsequently, the path encoding processing is performed on each layer leaf node in the expanded quadtree structure. Since the quadtree structure has been expanded, the result of path encoding will contain more information, making the encoding more complex and diverse. This complex encoding method makes it more difficult for attackers to infer the characteristics and patterns of the original data from the encoding result, thereby improving the security of the data. By using the above scheme, the application can flexibly adjust the structure of the quadtree as needed. By adding empty nodes, the application can adapt to different sizes and complexities of data sets, improving the adaptability and flexibility of data processing. Moreover, the expanded quadtree structure and the encoding result also provide a better basis for subsequent scrambling processing. Since the encoding result is more complex and diverse, the scrambling processing will be able to more effectively disrupt the order and association of the data, further improving the security of the data.

[0039] Optionally, after the step of encoding, the method further comprises:

[0040] Node updating: collecting real-time data of the current medical and health sensor, and obtaining a level to which the historical data of the current medical and health sensor belongs in the expanded quadtree structure, denoted as a first level; obtaining an empty node of the first level, denoted as a first node; filling the real-time data into the first node, taking a new encoding result corresponding to the first node as an encoding result of the real-time data, and updating the first node after adding the real-time data to a leaf node of the first level.

[0041] The application collects real-time data of the current medical and health sensor, and obtains a level to which the real-time data belongs in the quadtree structure, denoted as a first level. Subsequently, the application obtains an empty node in the first level, denoted as a first node. By obtaining the empty node, the application can find a suitable storage location for the real-time data while maintaining the integrity and consistency of the quadtree structure. Subsequently, the application fills the real-time data into the first node, and takes an encoding result of the first node as an encoding result of the real-time data, realizing dynamic integration of the real-time data and the quadtree structure, so that the real-time data can immediately participate in the subsequent data processing process. The first node after adding the real-time data is then updated to a leaf node of the first level, further enriching the content of the quadtree structure and increasing the complexity and diversity of the encoding. By adopting the above scheme, the application can obtain and process real-time data of the sensor in real time, maintaining the timeliness and accuracy of data processing. By filling the real-time data into the quadtree structure, the application can dynamically adjust the data structure to adapt to changes and updates of the data.

[0042] In a second aspect, the application provides a medical and health sensor data security protection system, which adopts the following technical scheme:

[0043] The medical and health sensor data security protection system comprises:

[0044] A historical data collection module is configured to collect historical data of the medical and health sensor, and obtain user information of the medical and health sensor.

[0045] A data matrix construction module is configured to extract a plurality of features of the user information, and encode each feature individually; construct a first matrix for each medical and health sensor, wherein diagonal elements of the first matrix are encoding values of any one feature, non-diagonal elements of the first matrix are historical data originating from the same medical and health sensor, and when the historical data is insufficient to fill the non-diagonal elements of the first matrix, a specific character is used to fill the non-diagonal elements of the first matrix, and the same label is added to the historical data belonging to the same first matrix; and a block splicing algorithm is used to splice the first matrix into a second matrix.

[0046] The data matrix splitting module is configured to split the second matrix by using a quadtree splitting algorithm to obtain a plurality of split blocks.

[0047] The attribution judgment module is configured to determine whether the historical data in the current split block belongs to the same first matrix according to the label.

[0048] The leaf node adding module is configured to integrate the historical data in the current split block as a leaf node in the quadtree structure.

[0049] The data matrix updating module is configured to update the current split block as the second matrix and trigger the data matrix splitting module.

[0050] The encoding module is configured to perform path encoding processing on the leaf node to obtain an encoding result.

[0051] The scrambling module is configured to perform scrambling processing on the leaf node by taking the encoding result as a label of the leaf node to obtain and store a processing result.

[0052] The application first obtains historical data and user information, converts features by the feature extraction module, forms a matrix and splices by the data matrix constructing module, and then splits the matrix into a plurality of blocks by the data matrix splitting module using a quadtree algorithm. After the attribution judgment module determines the consistency of data attribution, the leaf node adding module integrates the data into the quadtree structure, and the data matrix updating module is responsible for dynamic updating and cyclic processing. The encoding module performs path encoding on the leaf node to improve complexity, and finally the scrambling module enhances security by label scrambling. The modules are coordinated to realize efficient, flexible and secure data processing, and comprehensively guarantee the security of data in the medical care health sensor.

[0053] In a third aspect, the application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program, when processed by a processor, is configured to implement the method of the first aspect.

[0054] In a fourth aspect, the application provides a server having the system of the second aspect loaded thereon, wherein the server comprises a processor and a memory connected in communication with the processor.

[0055] The memory is provided with the computer readable storage medium of the third aspect, and the computer readable storage medium has a computer program stored thereon.

[0056] The processor processes the computer program stored on the computer readable storage medium to implement the method of the first aspect.

[0057] In summary, the application has at least one of the following beneficial technical effects:

[0058] 1.The application first extracts a plurality of features in the user information, encodes the features to obtain an encoded value, then constructs a first matrix using the encoded value, and constructs a second matrix through a block splicing algorithm. The application not only retains the original features of the data, but also increases the complexity by introducing diagonal elements. Through this disordering process, the security of the data is improved. Subsequently, the application uses a quadtree segmentation algorithm to segment the second matrix, so that each segmentation block contains a meaningful data subset. Subsequently, the application realizes dynamic recognition of data attribution by judging whether the historical data in the current segmentation block belongs to the same first matrix. When the data in the segmentation block belongs to the same first matrix, it is directly added as a leaf node; otherwise, the segmentation block is updated to a second matrix and continues to be segmented to realize the layer-by-layer refinement of data, while improving the efficiency and accuracy of data processing. Subsequently, the application performs path encoding processing on the leaf nodes, converting the complex data structure into an encoded result that is easy to manage and transmit, thereby simplifying the data processing process and increasing the security of the data. Finally, the application takes the encoded result as the label of the leaf node and performs shuffling processing on the leaf node, thereby breaking the original association between the data, making it difficult for attackers to infer the original data from the processing result, and significantly improving the security of the data.

[0059] 2.The application splices the first matrix in the order of the first correlation degree from large to small, so that the features with higher correlation are adjacent or close in the second matrix, thereby making the adjacent nodes of the quadtree structure have higher correlation, that is, in the quadtree construction process, the high-correlation features arranged adjacent or close will be preferentially divided into the same node or adjacent node. This spatial topology structure enables the quadtree to naturally form a feature clustering effect when it is split, that is, a feature subset with similar semantics or statistical correlation is automatically merged into the same subtree. When performing a region query or a nearest neighbor search, the search path can be predicted based on the correlation strength between nodes, thereby improving the retrieval efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0060] Fig. 1 is a flowchart of embodiment 1 of the application;

[0061] Fig. 2 is a flowchart of embodiment 2 of the application;

[0062] Fig. 3 is a flowchart of embodiment 3 of the application. DETAILED DESCRIPTION

[0063] The application will be further described in detail below. Figs. 1 to 3

[0064] Embodiment 1: This embodiment discloses a medical and health sensor data security protection method, referring to​Fig. 1 The method comprises: S11 historical data collection, S12 data matrix construction, S13 data matrix segmentation, S14 attribute judgment, S15 leaf node addition, S16 data matrix update, S17 encoding, and S18 scrambling. First, the historical data of the medical care health sensor is collected, and user information is obtained. Then, the features of the user information are extracted and encoded. A first matrix is constructed using the encoded values of any feature. A second matrix is constructed through a block splicing algorithm. Then, a quadtree segmentation algorithm is used to segment the second matrix, and the attribution of the segmented block data is judged. If the attribution is the same first matrix, the segmented block is taken as a leaf node of the quadtree structure. Otherwise, the second matrix is updated and segmentation is continued. Finally, the leaf nodes are path-encoded, and the encoded results are scrambled to obtain and store the final processing result. The process of this embodiment is as follows:

[0065] S11 historical data collection, the historical data of the medical care health sensor is collected, and the user information of the medical care health sensor is obtained.

[0066] The historical data includes various physiological parameters and environmental data recorded by the medical care health sensor.

[0067] Meanwhile, this embodiment also obtains user information associated with the medical care health sensor, such as user identity information, usage habits, and health status.

[0068] S12 construction of data matrix, multiple features of user information are extracted, and each feature is encoded separately to convert the features into encoded values that can be recognized and processed by a computer.

[0069] This embodiment constructs a first matrix for each medical care health sensor, and selects the encoded value of any feature as the diagonal element of the first matrix. The historical data of the same medical care health sensor is filled into the non-diagonal element of the first matrix. Then, this embodiment also adds the same label to the historical data belonging to the same first matrix. Then, this embodiment uses a block splicing algorithm to splice multiple first matrices into a larger second matrix to further increase the complexity and diversity of the data.

[0070] In this embodiment, if the historical data of the same medical care health sensor cannot form a complete first matrix, a specific character is used to fill in the missing positions in the first matrix.

[0071] In other embodiments, in order to facilitate calculation, a time window can be set according to the needs, and the historical data of the medical care health sensor within the time window is filled into the first matrix.

[0072] S13 data matrix segmentation, after the second matrix is constructed, a quadtree segmentation algorithm is used to segment it. The quadtree segmentation algorithm can recursively segment a large data block into multiple smaller sub-blocks until the second matrix is segmented into multiple segmented blocks with different characteristics and attributes, obtaining a quadtree structure.

[0073] S14 attribute judgment, after multiple segmented blocks are segmented, this step will judge whether the historical data in each segmented block belongs to the same first matrix according to the label added before. If the judgment result is yes, it means that the data in the current segmented block comes from the same first matrix, and then S15 leaf node addition will be performed; if the judgment result is no, it means that the data in the current segmented block comes from different first matrices, and then S16 data matrix update will be performed.

[0074] S15 leaf node addition, if the historical data in the current segmented block belongs to the same first matrix, this step will take the historical data in the current segmented block as a leaf node in the quadtree structure.

[0075] S16 data matrix update, if the historical data in the current segmented block does not belong to the same first matrix, this step will update the current segmented block to the second matrix, and re-perform S13 data matrix segmentation to continue segmenting and processing the current segmented block.

[0076] S17 encoding, after the quadtree structure is constructed, the leaf nodes in the quadtree structure are processed by path encoding. Path encoding is an encoding method that can uniquely identify the position of a leaf node in a quadtree, and this step can obtain the encoding result of each leaf node. These encoding results not only have uniqueness and recognizability, but also provide a basis for subsequent data scrambling processing.

[0077] S18 scrambling, taking the encoding result as the label of the leaf node, the leaf nodes in the quadtree structure are processed by scrambling. Scrambling is a processing method that can disrupt the order and association of leaf nodes, and this step can further improve the security and privacy of data through scrambling processing. The result after scrambling processing will be stored by the system.

[0078] After the user's identity is verified, this embodiment will feedback the processing result to the user.

[0079] The following is an example to illustrate this embodiment.

[0080] S11 historical data acquisition, historical data of medical and health sensors is collected in a medical and health monitoring project. These medical and health sensors are installed in the home of each user to monitor the health status of the user for a long time.

[0081] The collected historical data includes various physiological parameters recorded by the sensors, such as heart rate, blood pressure (high and low pressure concatenated), blood sugar level, etc., as well as environmental data, such as indoor temperature, humidity, air quality, etc.

[0082] Subsequently, user information associated with these sensors is obtained. For example, user identity information (age, gender, etc.), usage habits (sensor usage frequency, usage time period, etc.), and health status (whether or not having chronic diseases).

[0083] S12 constructs a data matrix, extracts various features from the user information, extracts usage frequency features from the usage habits, and extracts whether or not having high blood pressure features from the health status, etc. The extracted features are processed by encoding and converted into encoded values.

[0084] For example, the encoding method for each feature in this step is as follows:

[0085] Age: interval encoding is used. That is, the age is divided into different intervals, and each interval corresponds to a unique encoding value.

[0086] 0 - 18 years old: encoded as 01

[0087] 19 - 30 years old: encoded as 02

[0088] 31 - 45 years old: encoded as 03

[0089] 46 - 60 years old: encoded as 04

[0090] 60 years old and above: encoded as 05

[0091] Gender: binary encoding is used. Male is 0 and female is 1.

[0092] Usage frequency: hierarchical encoding is used.

[0093] More than 3 times a day: encoded as 13

[0094] More than 1 time and less than 3 times a week: encoded as 12

[0095] More than 1 time and less than 3 times a month: encoded as 11

[0096] More than 1 time and less than 3 times a year: encoded as 10

[0097] Usage time period: a day is divided into different time periods, and each time period corresponds to an encoding value.

[0098] 0:00 - 6:00: encoded as 21

[0099] 6:00 - 12:00: encoded as 22

[0100] 12:00 - 18:00: encoded as 23

[0101] 18:00 - 24:00: encoded as 24

[0102] Health status: encoded in binary, 30 means yes, 31 means no.

[0103] For example, there is a user information: 32 years old, male, more than 1 time and less than 3 times per week, 18:00-24:00 time period, no chronic disease, its encoding is: 03, 0, 11, 24, 30.

[0104] The non-diagonal elements of the first matrix are filled with historical data from the same medical and nursing health sensor. If the historical data of the same medical and nursing health sensor cannot form a complete first matrix, it is filled with a specific character (X in this embodiment).

[0105] For example, assuming that the encoding value of the "32 years old" feature is 03, 03 is taken as the diagonal element of the first matrix, and the heart rate data of Zhang San in a period of time is selected to fill the non-diagonal position:

[0106]

[0107] The elements in the above first matrix are marked as matrix one.

[0108] For example, assuming that the encoding value of the "gender" feature is 0, 0 is taken as the diagonal element of the first matrix, and the blood pressure data of Zhang San in a period of time is selected to fill the non-diagonal position:

[0109]

[0110] The elements in the above first matrix are marked as matrix two.

[0111] Then, the system uses a block splicing algorithm to splice multiple first matrices into a second matrix.

[0112]

[0113] S13 data matrix segmentation, after the second matrix is constructed, a quadtree segmentation algorithm is used to segment it. The quadtree segmentation algorithm can recursively divide a large data block into multiple smaller sub-blocks.

[0114] S14 attribute judgment, after the plurality of split blocks are segmented, whether the historical data in each split block belongs to the same first matrix is judged in turn by checking the mark of each data in the split block. If the judgment result is yes, it means that the data in the current split block has consistency, and then S15 leaf node addition will be performed; if the judgment result is no, it means that the data in the current split block comes from different first matrices, and then S16 data matrix update will be performed.

[0115] S15 leaf node addition, the historical data in the current split block is taken as a leaf node in the quadtree structure.

[0116] S16 data matrix update, if the historical data in the current split block does not belong to the same first matrix, the current split block is updated to the second matrix, and S13 data matrix segmentation is re-executed to continue the segmentation and processing of the updated second matrix.

[0117] S17 encoding, after the quadtree structure is constructed, the path encoding processing is performed on the leaf nodes in the quadtree structure. Path encoding is an encoding method that can uniquely identify the position of the leaf node in the quadtree.

[0118] For example, for a leaf node in the quadtree, according to the path from the root node to the leaf node, a unique code is generated, such as 0110, which means that the leaf node is reached by first going to the left subtree, then going to the right subtree, then going to the right subtree, and finally going to the left subtree from the root node.

[0119] S18 scrambling, the encoding result is taken as the label of the leaf node, and the leaf nodes in the quadtree structure are scrambled.

[0120] The embodiment first collects historical data and user information of the medical and health sensor, then extracts user information features and encodes, constructs a first matrix with the encoded value and splices it into a second matrix; then the quadtree segmentation algorithm is used to segment the second matrix, the data attribution of the split block is judged, if it belongs to the same first matrix, it is added to the quadtree structure as a leaf node, otherwise the split block is updated to the second matrix and the segmentation is continued; after the quadtree construction is completed, the path encoding is performed on the leaf nodes, and finally the encoding result is taken as the label to scramble the leaf nodes, and the processing result is stored to improve the security and privacy of the data.

[0121] Embodiment 2: refer to Fig. 2 The difference between this embodiment and embodiment 1 is that S12 constructing a data matrix further includes:

[0122] S21 quantity judgment, compare the number of extracted features with the number of medical and health sensors. If the number of features is less than the number of medical and health sensors, it means that the current encoding value is not enough to fill all the first matrix, and the diagonal elements of the first matrix in embodiment 1 repeat the encoding value of a certain or several features, in order to increase the difference, this embodiment executes S22 to increase the number of features; if the number of features is greater than or equal to the number of medical and health sensors, it means that the current features are already rich enough, and directly execute S13 data matrix segmentation.

[0123] S22 increase features, statistical analysis is performed on the behavior data of the user to mine more valuable new features. These new features include the activity frequency, use time, behavior mode and the like of the user, which can more comprehensively reflect the behavior habits and health status of the user.

[0124] The new features are merged with the features in S12 to construct the data matrix, and the cosine similarity algorithm is used to calculate the correlation between any two features, which is denoted as the first correlation.

[0125] Taking a 32-year-old person without chronic diseases as an example, the process of calculating the correlation between any two features by using the cosine similarity algorithm is as follows:

[0126] The above features are one-hot encoded (the vector length is the number of features, the element of the position corresponding to the feature is 0, and the remaining positions are 1), and the one-hot encoded vectors are [0, 1, 1, 1, 1] and [1, 1, 1, 1, 0] respectively. The correlation calculation process is as follows:

[0127] First, calculate the dot product of the two vectors:

[0128] 0x1+1x1+1x1+1x1+0x1=3;

[0129] Then calculate the length of the two vectors:

[0130] ;

[0131] ;

[0132] The first correlation is:

[0133] ;

[0134] The first correlation of other feature pairs is calculated in the same way as the above content, and this embodiment will not be described in more detail.

[0135] S23 historical data and feature correlation analysis, the cosine similarity algorithm is used to calculate the correlation between the historical data of each medical and health sensor and the ith feature, which is denoted as the second correlation.

[0136] The historical data of each medical and health sensor is first integrated into a vector, denoted as a first vector, and then the encoding value of the i-th feature is arranged into a vector with the same dimension as the first vector, denoted as a second vector, and then the cosine similarity between the first vector and the second vector, i.e., the second correlation, is calculated.

[0137] Taking the first vector [89, 78, 74, 82, 71] as an example, the second vector is [03, 03, 03, 03, 03], and then the cosine similarity algorithm is used to calculate the second correlation of the first vector and the second vector.

[0138] In other embodiments, the same feature as S22 can also be selected, and the historical data with the same dimension as the second vector [0, 1, 1, 1, 1] is combined to form the first vector to calculate the second correlation.

[0139] Then, S13 data matrix segmentation is performed. When performing S13 data matrix segmentation, first, the historical data of each medical and health sensor is filled into the corresponding first matrix according to the order of the second correlation from large to small (for example, matrix one uses the gender feature as the diagonal line element, and then this embodiment fills the blood pressure into matrix one according to the order of the second correlation from large to small, and matrix two uses the use time period feature as the diagonal line element, and then this embodiment fills the medical and health sensor with the largest correlation with the use time period into matrix two), and then the first matrix is spliced according to the order of the first correlation from large to small.

[0140] S24 data analysis, set the normal range of the output data of each medical and health sensor. These normal ranges are determined based on historical data, industry standards or medical knowledge, for example, the normal range of heart rate is 60-105.

[0141] S25 range judgment, sequentially judges whether the j-th non-diagonal line element in the first matrix falls within the normal range, if yes, it means that the data of the element is normal, and the j-th non-diagonal line element in the first matrix is retained, and S26 iteration is performed; if not, it means that the data of the element may be abnormal, and the j-th non-diagonal line element is replaced with a specific character.

[0142] S26 iteration, the j+1-th non-diagonal line element is taken as the new j-th non-diagonal line element, and S25 range judgment is performed, until all non-diagonal line elements in the first matrix are traversed.

[0143] The embodiment firstly determines whether the feature quantity is sufficient, and if not, carries out statistical analysis of user behavior data to increase the features, and integrates the new and old features to form a new feature set; then the cosine similarity is used to calculate the correlation between the new features and the correlation between each medical and nursing health sensor historical data and the new features, and then the historical data is sequentially filled into the first matrix according to the second correlation, and is spliced into the second matrix according to the first correlation; then the normal range of the output data of each sensor is set, and it is checked whether the non-diagonal elements of the first matrix are within the normal range through the range judgment step, if yes, the element is retained and the judgment is continued, otherwise the abnormal element is replaced with a specific character, until the preset stop condition is met, thereby improving the data processing quality and safety.

[0144] Embodiment 3: Refer to Fig. 3 The difference between the embodiment and the embodiment 1 is that after performing the attribute judgment of S14, before performing the data matrix updating of S16, it further includes:

[0145] S31 segmentation block judgment, judging whether the quantity of historical data contained in the current segmentation block is less than a preset minimum quantity threshold, if yes, performing the data matrix updating of S16; if no, performing the segmentation block updating of S32.

[0146] S32 segmentation block updating, calculating the covariance matrix of the non-diagonal elements in the current segmentation block, and calculating the maximum eigenvalue of the covariance matrix and the corresponding eigenvector. The covariance matrix can reflect the correlation and distribution characteristics between the data, and the maximum eigenvalue and the corresponding eigenvector can reveal the main change direction and amplitude of the data.

[0147] According to the direction of the eigenvector, the current segmentation block is segmented to obtain two smaller sub-blocks, i.e. the first segmentation sub-block and the second segmentation sub-block. This step can refine the data in the main change direction, i.e. according to the direction of the maximum eigenvector to segment the first segmentation sub-block and the second segmentation sub-block to maximize the difference between them.

[0148] The first segmentation sub-block and the second segmentation sub-block are sequentially updated as the current segmentation block, and the data matrix updating of S16 is performed.

[0149] After performing the data matrix updating of S16, before performing the encoding of S17, it further includes:

[0150] S33 increase empty nodes, obtain the current quadtree structure, and add multiple empty nodes to the leaf nodes of the kth layer. These empty nodes will be updated to the leaf nodes of the kth layer subsequently, thereby forming an expanded quadtree structure to increase the depth and complexity of the quadtree.

[0151] The encoding of S17 performs path encoding processing on each layer leaf node in the expanded quadtree structure.

[0152] After performing S17 encoding, before performing S18 scrambling, further comprising:

[0153] S34 node updating, collecting real-time data of the current medical and health sensor, and obtaining a level to which historical data of the current medical and health sensor belongs in the quadtree structure, denoted as a first level, to associate the real-time data with a specific level in the quadtree structure.

[0154] Obtaining an empty node in the first level, denoted as a first node. Filling the real-time data to the first node, taking the encoding result of the first node as the encoding result of the real-time data, updating the first node after adding the real-time data to a leaf node of the first level, and integrating the real-time data into the quadtree structure to realize dynamic processing and encoding of the real-time data.

[0155] The embodiment first judges whether the number of historical data contained in the current segmentation block is less than a preset threshold, and if yes, the step of data matrix updating is performed, and if not, the covariance matrix and the maximum eigenvalue and the corresponding eigenvector thereof are calculated, and the current segmentation block is updated in turn after being divided into two sub-blocks in the direction of the eigenvector for continuous processing; subsequently, the quadtree structure is expanded to increase empty nodes; path encoding is performed on the leaf nodes of each layer after expansion during encoding; after encoding and before scrambling, real-time data is collected and associated with the first level of the quadtree, and the empty node of the level is obtained to fill the real-time data, and the encoding result thereof is taken as the encoding of the real-time data, and the node is updated to a leaf node, so as to realize dynamic processing, encoding and structure expansion of data.

[0156] Embodiment 4: The embodiment discloses a medical and health sensor data security protection system, the system comprising:

[0157] A historical data collection module is responsible for collecting historical data of medical and health sensors and obtaining user information associated with these sensors.

[0158] A data matrix construction module is responsible for extracting features from user information, which can highly summarize and represent certain characteristics or behavior patterns of users, and then encoding the features to obtain encoding values.

[0159] The module is also used to construct a first matrix with the encoding value of any feature as a diagonal element, and the non-diagonal element of the first matrix is historical data from the same medical and health sensor. When the historical data is insufficient to fill the non-diagonal element of the first matrix, a specific character is used to fill the non-diagonal element of the first matrix, and a mark is added to the data belonging to the same first matrix. The module is also used to splice the first matrix into a second matrix using a block splicing algorithm.

[0160] The data matrix division module divides the second matrix into multiple division blocks with different characteristics and attributes by using a quadtree division algorithm after the second matrix is constructed.

[0161] The attribution judgment module judges whether the historical data in the current division block belongs to the same first matrix according to the mark.

[0162] The leaf node adding module adds the historical data in the current division block as a leaf node in the quadtree structure.

[0163] The data matrix updating module updates the current division block as the second matrix and triggers the data matrix division module.

[0164] The encoding module encodes the leaf node to obtain an encoding result.

[0165] The scrambling module scrambles the encoding result as a label of the leaf node and performs scrambling processing on the leaf node in the quadtree structure. The result after the scrambling processing is stored by the system for subsequent data analysis and application.

[0166] Embodiment 5: The embodiment provides a computer readable storage medium, and a computer program is stored on the computer readable storage medium. The computer program is used to implement the method when the computer program is processed by a processor.

[0167] Embodiment 6: The embodiment provides a server, and the server loads the system. The server comprises a processor and a memory connected with the processor in communication.

[0168] The memory is provided with the computer readable storage medium, and the computer readable storage medium stores the computer program.

[0169] The processor processes the computer program stored on the computer readable storage medium to implement the method.

[0170] The above are preferred embodiments of the application, and do not limit the protection scope of the application. Any equivalent changes made according to the structure, shape and principle of the application should be covered within the protection scope of the application.

Claims

1. A medical and health sensor data security protection method, characterized in that, Comprise: S11 historical data collection: collect the historical data of the medical care health sensor, obtain the user information of the medical care health sensor, and perform the step of S12 constructing a data matrix; S12 Constructing a data matrix: extracting a plurality of features of user information, and separately encoding each feature; A first matrix is constructed for each medical care health sensor, the diagonal elements of the first matrix are the encoding values of any one feature, the non-diagonal elements of the first matrix are the historical data from the same medical care health sensor, and when the historical data is insufficient to fill the non-diagonal elements of the first matrix, a specific character X is used to fill the non-diagonal elements of the first matrix, and the same label is added to the historical data belonging to the same first matrix; using a block splicing algorithm, the first matrix is spliced into a second matrix, and the step of S13 data matrix segmentation is performed; S13 Data matrix segmentation: using a quadtree segmentation algorithm to segment the second matrix to obtain a plurality of segmentation blocks, and performing the step of S14 attribute judgment; S14 Attribute judgment: judging whether the historical data in the current segmentation block belongs to the same first matrix according to the label, if yes, performing the step of S15 leaf node addition; If not, S16 performs the step of data matrix updating; S15 Leaf node addition: the historical data in the current segmentation block is taken as a leaf node in the quadtree structure, and the step of S17 encoding is performed; S16 Data matrix updating: updating the current segmentation block to the second matrix, and performing the step of S13 data matrix segmentation; S17 Encoding: performing path encoding processing on the leaf node to obtain an encoding result, and performing the step of S18 shuffling; S18 Shuffling: taking the encoding result as the label of the leaf node, performing shuffling processing on the leaf node, and obtaining and storing the processing result; The step of S12 constructing a data matrix further comprises: S21 Quantity judgment: judging whether the number of features is less than the number of medical care health sensors, if yes, performing the step of S22 increasing features; If not, directly perform the step of S13 constructing a data matrix; S22 Increase features: statistically analyzing the behavior data of the user to obtain new features, merging the new features with the features in the step of S12 constructing a data matrix, and using a cosine similarity algorithm to calculate the correlation between any two features, denoted as a first correlation; In the step of S12 constructing a data matrix, the first matrix is spliced in the order of the first correlation from large to small.

2. The medical and health sensor data security protection method according to claim 1, characterized in that, After performing the step of S22 increasing features, the method further comprises: S23 Historical data and feature correlation analysis: using a cosine similarity algorithm to calculate the correlation between the historical data of each medical care health sensor and the i-th feature, denoted as a second correlation; In the step of S12 constructing a data matrix, the historical data of each medical care health sensor is filled into the corresponding first matrix in the order of the second correlation from large to small.

3. The medical and health sensor data security protection method according to claim 2, characterized in that, The step of S12 constructing a data matrix further comprises: S24 Data analysis: setting the normal range of the output data of each medical care health sensor, and performing the step of S25 range judgment; S25 range judgment: whether the jth non-diagonal element in each first matrix falls within the normal range is judged in turn, if yes, the jth non-diagonal element in the first matrix is retained, and the steps of S26 iteration are executed; if not, the jth non-diagonal element is replaced by a specific character X; S26 iteration: the j+1th non-diagonal element is taken as the new jth non-diagonal element, and the steps of S25 range judgment are executed until all non-diagonal elements in the first matrix are traversed.

4. The medical and health sensor data security protection method according to any one of claims 1-3, characterized in that, After the step of S14 attribute judgment is executed, before the step of S16 data matrix update is executed, the method further comprises: S31 segmentation block judgment: whether the number of historical data contained in the current segmentation block is less than the preset minimum number threshold is judged, if yes, the step of S16 data matrix update is executed; if not, the step of S32 segmentation block update is executed; S32 segmentation block update: the covariance matrix of the non-diagonal elements in the current segmentation block is calculated, the maximum eigenvalue of the covariance matrix and the corresponding eigenvector are calculated, the current segmentation block is segmented according to the direction of the eigenvector to obtain a first segmentation sub-block and a second segmentation sub-block, the first segmentation sub-block and the second segmentation sub-block are updated to the current segmentation block in turn, and the step of S16 data matrix update is executed.

5. The medical and health sensor data security protection method according to claim 4, characterized in that, After the step of S16 data matrix update is executed, the method further comprises: S33 increase empty node: obtain a quadtree structure, increase a plurality of empty nodes for the kth layer leaf node in the quadtree structure, update the increased empty nodes to the kth layer leaf node in the quadtree structure, obtain an expanded quadtree structure, and execute the step of S17 coding; In the step of S17 coding, each layer leaf node in the expanded quadtree structure is subjected to path coding processing to obtain a new coding result.

6. The medical and health sensor data security protection method according to claim 5, characterized in that, After the step of S17 coding is executed, the method further comprises: S34 node update: collecting real-time data of the current medical care health sensor, and obtaining the level to which the historical data of the current medical care health sensor belongs in the expanded quadtree structure, denoted as the first level; obtaining the empty node of the first level, denoted as the first node; filling the real-time data into the first node, taking the new coding result corresponding to the first node as the coding result of the real-time data, and updating the first node after adding the real-time data to the leaf node of the first level.

7. A medical and wellness sensor data security protection system for implementing the method according to any one of claims 1-6, characterized by, Comprise: a historical data acquisition module for acquiring historical data of a medical care health sensor and obtaining user information of the medical care health sensor; a data matrix construction module for extracting a plurality of features of the user information and individually coding each feature; a first matrix is constructed for each medical care health sensor, the diagonal elements of the first matrix are the coding values of any one feature, the non-diagonal elements of the first matrix are the historical data from the same medical care health sensor, and when the historical data is insufficient to fill the non-diagonal elements of the first matrix, the non-diagonal elements of the first matrix are filled with a specific character X, and the historical data belonging to the same first matrix is marked with the same mark; the first matrix is spliced into a second matrix by using a block splicing algorithm; The data matrix splitting module is configured to split the second matrix by using a quadtree splitting algorithm to obtain a plurality of split blocks. The attribution judgment module is configured to determine whether the historical data in the current split block belongs to the same first matrix according to the mark. The leaf node adding module is configured to add the historical data in the current split block as a leaf node in the quadtree structure. The data matrix updating module is configured to update the current split block as the second matrix and trigger the data matrix splitting module. The encoding module is configured to perform path encoding processing on the leaf node to obtain an encoding result. The scrambling module is configured to perform scrambling processing on the leaf node by taking the encoding result as a label of the leaf node to obtain and store a processing result.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that The computer program, when processed by the processor, is configured to implement the method according to any one of claims 1-6.

9. A server, characterized by The server loads the system according to claim 7, and the server comprises a processor and a memory connected with the processor in communication. The memory is provided with the computer readable storage medium according to claim 8, and the computer readable storage medium stores a computer program. When the processor processes the computer program stored in the computer readable storage medium, the method according to any one of claims 1-6 is implemented.

Images