Doctor-doctor information processing and analyzing system
The Yishitong information processing and analysis system generates standardized electronic medical records and data indexes through data collection, preprocessing, storage and security units, solving the data integration and security issues of medical information systems and realizing efficient data analysis and secure sharing.
Patent Information
- Application Number
- CN202510805579.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-17
- Publication Date
- 2025-09-26
AI Technical Summary
The existing medical information processing system has problems such as difficulty in data integration, insufficient analytical capabilities and imperfect information security, which leads to the inability to share and uniformly process information between different medical institutions, the inability to efficiently integrate data, insufficient analytical capabilities, and easy leakage of private data.
The Yishitong information processing and analysis system is used, including data collection, preprocessing, storage and security units. Through data preprocessing, valid data is determined, standardized electronic medical records and data indexes are generated, database associations are constructed, and dynamic encryption processing is performed in the cloud to achieve efficient data integration and secure sharing.
It improves the quality and standardization of medical data, increases data analysis efficiency, enhances data security, makes it easier for users to quickly obtain the information they need, and solves the problems of difficult data integration and imperfect security.
Smart Images
Figure CN120708928A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical information processing technology, and more particularly to a medical information processing and analysis system. Background Art
[0002] Currently, with the continuous development of information technology in the medical field, a vast amount of medical data, such as electronic medical records, imaging data, and test reports, is continuously generated. This data contains a wealth of information and is of great value for disease diagnosis, treatment plan development, and medical research. However, existing medical information processing systems suffer from the following problems: data integration is difficult, and incompatibility between information systems of different medical institutions prevents data sharing and unified processing; insufficient data analysis capabilities make it difficult to extract valuable information from massive amounts of medical data; and inadequate information security guarantees easily lead to the leakage of patients' private data.
[0003] Therefore, there is an urgent need for a system that can efficiently integrate, process and analyze medical data while ensuring information security. Summary of the Invention
[0004] In view of this, the present invention provides a Yishitong information processing and analysis system to solve the problems existing in the background technology.
[0005] In order to achieve the above object, the present invention adopts the following technical solutions:
[0006] A Yishitong information processing and analysis system includes: a data acquisition unit, a data preprocessing unit, a data analysis unit, a data storage unit, and a data security unit;
[0007] The data acquisition unit is used to collect original patient medical data from the hospital;
[0008] The data preprocessing unit performs credibility-based preprocessing on the original patient medical data to determine valid patient medical data;
[0009] The data storage unit is used to perform standardization processing operations on valid patient medical data and store the standardized patient medical data as a standard electronic medical record;
[0010] The data analysis unit generates a corresponding standard data table based on the standard electronic medical record, wherein the standard data table includes a table structure and table data; extracts patient identification information contained in the table data, and classifies and analyzes the table structure according to the patient identification information to generate a corresponding data index; stores the table data in a preset database, and establishes an association relationship between the preset database and the data index;
[0011] The data security unit uploads the preset database and the data index to the preset cloud server at the same time based on the association relationship, and dynamically encrypts the data storage unit so that the user can obtain the required patient medical information through the preset cloud server.
[0012] Optionally, the credibility-based preprocessing includes: taking the data change trend as the first credibility feature and the data vector feature as the second credibility feature, and determining the data credibility interval based on the first credibility feature and the second credibility feature; reading and calling the retrieved data, performing data reliability verification and compensation in combination with the credibility interval, and determining the valid patient medical data; wherein, the data reliability verification includes a first over-limit threshold, a second over-limit threshold and a third over-limit threshold, and if the first over-limit threshold is met, over-limit zeroing processing is performed; if the second over-limit threshold is met, interpolation completion processing is performed, wherein the interpolation completion method includes mean interpolation and trend interpolation; if the third over-limit threshold is met, re-collection compensation processing is performed.
[0013] Optionally, the data storage unit uses a normalization formula to normalize the diagnosis and treatment data. The data storage unit is provided with a normalization formula as follows:
[0014] F A =(f a -μ a ) / σ a
[0015] Among them, F A represents the standardized diagnosis and treatment data, f a represents diagnosis and treatment data, a represents diagnosis and treatment category, a∈A, μ a represents the average value of the diagnosis and treatment category in all normalized diagnosis and treatment data, σ a Represents the standard deviation of the diagnosis and treatment category in all normalized diagnosis and treatment data.
[0016] Optionally, the step of generating a corresponding standard data table based on the standard electronic medical record includes:
[0017] Performing a full scan of the standardized patient medical data to extract a first value corresponding to the indicator name, a second value corresponding to the indicator data, and a third value corresponding to the indicator result from the standardized patient medical data; mapping the first value, the second value, and the third value to preset locations corresponding to the indicator name, the indicator data, and the indicator result, respectively, and generating a corresponding indicator data chain based on the indicator name, the indicator data, and the indicator result, so as to generate the standard electronic medical record accordingly;
[0018] When a standard electronic medical record is obtained, a corresponding table header is generated according to the indicator data chain, and a table frame is generated surrounding the table header, the indicator data chain, and the first value, the second value, and the third value; based on the table frame, a first dividing line is generated between the indicator data chain and the table header, a second dividing line is generated between the indicator data chain and the first value, the second value, and the third value, and a third dividing line is generated between the first value and the second value and between the second value and the third value, so as to generate the corresponding standard data table, and the first dividing line, the second dividing line, and the third dividing line are all straight lines.
[0019] Optionally, the step of dynamically encrypting the preset database includes: when receiving a data access instruction sent by a user, parsing the data access instruction to extract the name of the accessing person and the access time carried in the data access instruction, and extracting the target letters contained in the name of the accessing person, the first number contained in the access time, and the second number contained in the table data;
[0020] A corresponding encryption key is randomly generated according to the target letter, the first number and the second number, so that the user can access the preset database through the encryption key, and the encryption key is unique.
[0021] Optionally, the step of randomly generating a corresponding encryption key based on the target letter, the first number and the second number includes: randomly arranging and combining the target letter, the first number and the second number to generate a number of original sequence passwords with preset bits, and screening the number of original sequence passwords to generate a number of corresponding target sequence passwords; and randomly selecting one of the target sequence passwords as the encryption key.
[0022] Optionally, the step of screening the plurality of original sequence codes includes:
[0023] detecting one by one the first number corresponding to the numbers and the second number corresponding to the letters in each of the original sequence codes, and determining in real time whether the first number is less than the second number;
[0024] If it is determined in real time that the first number is smaller than the second number, the original sequence code corresponding to the current first number being smaller than the second number is set as the target sequence code.
[0025] Optionally, the classification analysis is specifically:
[0026] Extracting semantic features of patient identification information; constructing a loss function based on the semantic features, and continuously optimizing the loss function to obtain a text information classification model;
[0027] The table architecture is classified and parsed using a text information classification model in combination with patient identification information.
[0028] Optionally, the semantic features of the patient identification information are extracted as follows:
[0029] Use the bag-of-words model to convert patient identification information text into word vectors;
[0030] The word vectors are sequentially input into the convolutional layer and the pooling layer to extract the structural features of the table architecture;
[0031] Segment the patient identification information text and calculate the position weight based on the position of the word in the table structure;
[0032] Calculate the TF-IDF value of each word;
[0033] Based on the Gaussian distribution principle, the TF-IDF value and position weight are integrated to calculate the comprehensive weight of the word and obtain the local features;
[0034] The structural features are fused with the local features to obtain the semantic features of the patient identification information;
[0035] Construct a loss function based on semantic features and obtain a table architecture classification model after optimization;
[0036] Use the classification model to classify the table schema and generate data indexes.
[0037] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a Yishitong information processing and analysis system, which includes a data acquisition unit, a data preprocessing unit, a data analysis unit, a data storage unit, and a data security unit. The data acquisition unit collects original patient medical data, the data preprocessing unit determines valid data based on credibility preprocessing, the data storage unit normalizes the valid data and stores it as a standard electronic medical record, the data analysis unit generates a standard data table based on the standard electronic medical record, extracts patient identification information and classifies and parses the table structure to generate a data index, stores the table data and establishes an association relationship between a preset database and the data index, and the data security unit uploads the preset database and data index to the cloud server based on the association relationship and dynamically encrypts the data storage unit. The present invention improves the quality and standardization of medical data, ensuring that the data is accurate and available; generates standardized data tables and performs classification and analysis to generate data indexes, which facilitates efficient management and rapid retrieval of data, improves data analysis efficiency, and helps to extract valuable information from massive data; in terms of data security, dynamic encryption processing and cloud storage mechanisms enhance data security and privacy protection, reduce the risk of data leakage, and at the same time facilitate users to obtain the required medical information through cloud servers, improve the convenience of data sharing and use, and help solve the problems of data integration difficulties, insufficient analysis capabilities and imperfect security guarantees in existing medical information processing systems. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.
[0039] Figure 1 This is a schematic diagram of the system structure provided by the present invention. DETAILED DESCRIPTION
[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0041] The embodiment of the present invention discloses a medical information processing and analysis system, such as Figure 1 As shown, it includes: a data acquisition unit, a data preprocessing unit, a data analysis unit, a data storage unit, and a data security unit;
[0042] The data collection unit is used to collect the hospital's original patient medical data; medical data includes patient data, diagnosis and treatment data, and medical information sources. Patient data includes patient name, patient ID number, and patient contact information. Diagnosis and treatment data includes diagnosis and treatment categories and contents. Diagnosis and treatment categories include time, medical items, and drugs. Diagnosis and treatment contents include consultation time, medical item test results, and doctor's medication selection. Medical information sources include doctor interactive input and patient interactive upload.
[0043] The data preprocessing unit performs credibility-based preprocessing on the original patient medical data to determine the valid patient medical data;
[0044] A data storage unit is used to perform standardized processing operations on valid patient medical data and store the standardized patient medical data as a standard electronic medical record;
[0045] The data analysis unit generates a corresponding standard data table based on the standard electronic medical record, the standard data table including a table structure and table data; extracts the patient identification information contained in the table data, and classifies and analyzes the table structure according to the patient identification information to generate a corresponding data index; stores the table data in a preset database, and establishes an association relationship between the preset database and the data index;
[0046] The data security unit uploads the preset database and data index to the preset cloud server at the same time based on the association relationship, and dynamically encrypts the data storage unit so that users can obtain the required patient medical information through the preset cloud server.
[0047] In a specific embodiment, credibility-based preprocessing includes: taking data change trend as the first credibility feature, taking data vector feature as the second credibility feature, and determining the data credibility interval based on the first credibility feature and the second credibility feature; reading and calling retrieval data, performing data reliability verification and compensation in combination with the credibility interval, and determining valid patient medical data; wherein, the data reliability verification includes a first over-limit threshold, a second over-limit threshold and a third over-limit threshold, and if the first over-limit threshold is met, over-limit zeroing processing is performed; if the second over-limit threshold is met, interpolation completion processing is performed, wherein the interpolation completion method includes mean interpolation and trend interpolation; if the third over-limit threshold is met, re-collection completion processing is performed.
[0048] In a specific embodiment, the storage unit uses a normalization formula to normalize the diagnosis and treatment data. The storage unit has a normalization formula as follows:
[0049] F A =(f a -μ a ) / σ a
[0050] Among them, F A represents the standardized diagnosis and treatment data, f a represents diagnosis and treatment data, a represents diagnosis and treatment category, a∈A, μ a represents the average value of the diagnosis and treatment category in all normalized diagnosis and treatment data, σ a Represents the standard deviation of the diagnosis and treatment category in all normalized diagnosis and treatment data.
[0051] In a specific embodiment, the steps of generating a corresponding standard data table according to a standard electronic medical record include:
[0052] Perform a full scan of the standardized patient medical data to extract a first value corresponding to the indicator name, a second value corresponding to the indicator data, and a third value corresponding to the indicator result from the standardized patient medical data; map the first value, the second value, and the third value to the preset positions corresponding to the indicator name, the indicator data, and the indicator result, respectively, and generate a corresponding indicator data chain based on the indicator name, the indicator data, and the indicator result to generate a standard electronic medical record; specifically, for example, the first value may be "gastroscopy 2.0 detection", the second number may be "stomach pH value", and the indicator result may be "PH between 6.1-6.6", and further, the above-mentioned first value, second value, and third value are respectively set directly below the above-mentioned indicator name, indicator data, and indicator result. At the same time, a corresponding indicator data chain is also generated in real time based on the above-mentioned indicator name, the above-mentioned indicator data, and the above-mentioned indicator result to generate a standard electronic medical record;
[0053] When a standard electronic medical record is obtained, a corresponding table header is generated according to the indicator data chain, and a table frame is generated to surround the table header, the indicator data chain, and the first value, the second value, and the third value; based on the table frame, a first dividing line is generated between the indicator data chain and the table header, a second dividing line is generated between the indicator data chain and the first value, the second value, and the third value, and a third dividing line is generated between the first value and the second value and between the second value and the third value, so as to generate a corresponding standard data table, and the first dividing line, the second dividing line, and the third dividing line are all straight lines.
[0054] It should be noted that after obtaining the required indicator data chain, table header, and first, second, and third values through the above steps, the above three types of data are taken as a whole, and a table frame that can surround the whole is further created, wherein the table header is at the top of the current table frame, the indicator data chain is in the middle, and the first, second, and third values are at the bottom.
[0055] In a specific embodiment, the step of dynamically encrypting a preset database includes: when receiving a data access instruction sent by a user, parsing the data access instruction to extract the name of the accessing person and the access time carried in the data access instruction, and extracting the target letters contained in the name of the accessing person, the first number contained in the access time, and the second number contained in the table data;
[0056] A corresponding encryption key is randomly generated according to the target letter, the first number and the second number, so that the user can access the preset database through the encryption key, and the encryption key is unique.
[0057] In a specific embodiment, the step of randomly generating a corresponding encryption key based on a target letter, a first number, and a second number includes: randomly permuting and combining the target letter, the first number, and the second number to generate a number of original sequence passwords with preset digits, and screening the number of original sequence passwords to generate a number of corresponding target sequence passwords; and randomly selecting a target sequence password as the encryption key.
[0058] It should be noted that by randomly permuting and combining the target letters, the first digit, and the second digit in real time, a number of original sequence passwords with a preset number of digits can be preliminarily generated. Preferably, the original sequence password consists of six digits. Specifically, for example, if the target letters obtained are "zhsan", the first digit obtained can be "0606", and the second digit obtained can be "1016". Further, the randomly generated original sequence passwords can be "zhsan6", "zhs060", "061016", and "zhsa10", etc. Furthermore, in order to further enhance the confidentiality of the encryption key, the current number of original sequence passwords will be further screened to further screen out a number of target sequence passwords. Based on this, only one target sequence password needs to be randomly selected to serve as the required encryption key.
[0059] In a specific embodiment, the step of screening a plurality of original sequence codes includes:
[0060] Detecting the first number corresponding to the numbers and the second number corresponding to the letters in each original sequence password one by one, and determining in real time whether the first number is less than the second number;
[0061] If it is determined in real time that the first number is smaller than the second number, the original sequence code corresponding to the current first number being smaller than the second number is set as the target sequence code.
[0062] In a specific embodiment, the classification analysis is specifically as follows:
[0063] Extract the semantic features of patient identification information; construct a loss function based on the semantic features, and continuously optimize the loss function to obtain a text information classification model;
[0064] Use the text information classification model and combine it with patient identification information to classify and parse the table structure.
[0065] In a specific embodiment, the semantic features of the patient identification information are extracted as follows:
[0066] Use the bag-of-words model to convert patient identification information text into word vectors;
[0067] The word vectors are sequentially input into the convolutional layer and the pooling layer to extract the structural features of the table architecture;
[0068] In an embodiment of the present invention, the word vector can be sequentially input into the convolution layer and the pooling layer to obtain structural features; wherein, the convolution layer feature extraction formula is:
[0069] c i =f(ω·H+b)
[0070] In the formula, f is the activation function, ω is the convolution kernel, b is the bias, H is the word vector, c is the word vector, i is the i-th feature vector extracted using the convolutional layer. The pooling layer uses the global maximum pooling (1-Max) process.
[0071] Segment the patient identification information text and calculate the position weight based on the position of the word in the table structure; you can use:
[0072]
[0073] Determine the position weight of the word; where ξ represents the preset parameter, f i represents the first occurrence position of word i in the corresponding text, η j Indicates the total number of words in text j, l i Indicates the position where word i last appears in the corresponding text.
[0074] Calculate the TF-IDF value of each word;
[0075] Based on the Gaussian distribution principle, the TF-IDF value and position weight are integrated to calculate the comprehensive weight of the word and obtain the local features;
[0076] The formula for determining the comprehensive weight is:
[0077]
[0078] LTFIDF(i,j)=TF-IDF(i,j)×Eigenvalue(i,j)
[0079] Where LTFIDF(i,j) represents the comprehensive weight of the i-th word in document j, TF-IDF(i,j) represents the TF-IDF value of word i, σ represents the variance of the Gaussian distribution, and μ represents the expectation of the Gaussian distribution;
[0080] The comprehensive weight of the word is multiplied by the word vector of the corresponding word to obtain the local feature.
[0081] The structural features are fused with the local features to obtain the semantic features of the patient identification information;
[0082] Among them, the language feature fusion formula is:
[0083]
[0084] In the above formula, O represents language features, G represents structural features, and M represents local features.
[0085] Construct a loss function based on semantic features and obtain a table architecture classification model after optimization;
[0086] Use the classification model to classify the table schema and generate data indexes.
[0087] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.
[0088] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A Yishitong information processing and analysis system, characterized in that: include: Data acquisition unit, data preprocessing unit, data analysis unit, data storage unit, data security unit; The data acquisition unit is used to collect original patient medical data from the hospital; The data preprocessing unit performs credibility-based preprocessing on the original patient medical data to determine valid patient medical data; The data storage unit is used to perform standardization processing operations on valid patient medical data and store the standardized patient medical data as a standard electronic medical record; The data analysis unit generates a corresponding standard data table based on the standard electronic medical record, wherein the standard data table includes a table structure and table data; extracts patient identification information contained in the table data, and classifies and analyzes the table structure according to the patient identification information to generate a corresponding data index; stores the table data in a preset database, and establishes an association relationship between the preset database and the data index; The data security unit uploads the preset database and the data index to the preset cloud server at the same time based on the association relationship, and dynamically encrypts the data storage unit so that the user can obtain the required patient medical information through the preset cloud server.
2. The Yishitong information processing and analysis system according to claim 1, characterized in that: The credibility-based preprocessing includes: taking the data change trend as the first credibility feature and the data vector feature as the second credibility feature, and determining the data credibility interval based on the first credibility feature and the second credibility feature; reading and calling the retrieved data, performing data reliability verification and compensation in combination with the credibility interval, and determining the valid patient medical data; wherein, the data reliability verification includes a first over-limit threshold, a second over-limit threshold and a third over-limit threshold, and if the first over-limit threshold is met, over-limit zeroing processing is performed; if the second over-limit threshold is met, interpolation completion processing is performed, wherein the interpolation completion method includes mean interpolation and trend interpolation; if the third over-limit threshold is met, re-collection compensation processing is performed.
3. The Yishitong information processing and analysis system according to claim 1, characterized in that: The data storage unit uses a normalization formula to normalize the diagnosis and treatment data. The data storage unit is provided with a normalization formula as follows: F A =(f a -m a ) / s a Among them, F A represents the standardized diagnosis and treatment data, f a represents diagnosis and treatment data, a represents diagnosis and treatment category, a∈A, μ a represents the average value of the diagnosis and treatment category in all normalized diagnosis and treatment data, σ a Represents the standard deviation of the diagnosis and treatment category in all normalized diagnosis and treatment data.
4. The Yishitong information processing and analysis system according to claim 1, characterized in that: The step of generating a corresponding standard data table according to the standard electronic medical record includes: Performing a full scan of the standardized patient medical data to extract a first value corresponding to the indicator name, a second value corresponding to the indicator data, and a third value corresponding to the indicator result from the standardized patient medical data; mapping the first value, the second value, and the third value to preset locations corresponding to the indicator name, the indicator data, and the indicator result, respectively, and generating a corresponding indicator data chain based on the indicator name, the indicator data, and the indicator result, so as to generate the standard electronic medical record accordingly; When a standard electronic medical record is obtained, a corresponding table header is generated according to the indicator data chain, and a table frame is generated surrounding the table header, the indicator data chain, and the first value, the second value, and the third value; based on the table frame, a first dividing line is generated between the indicator data chain and the table header, a second dividing line is generated between the indicator data chain and the first value, the second value, and the third value, and a third dividing line is generated between the first value and the second value and between the second value and the third value, so as to generate the corresponding standard data table, and the first dividing line, the second dividing line, and the third dividing line are all straight lines.
5. The Yishitong information processing and analysis system according to claim 1, characterized in that: The step of dynamically encrypting the preset database includes: when receiving a data access instruction sent by a user, parsing the data access instruction to extract the name of the access person and the access time carried in the data access instruction, and extracting the target letter contained in the name of the access person, the first number contained in the access time, and the second number contained in the table data; A corresponding encryption key is randomly generated according to the target letter, the first number and the second number, so that the user can access the preset database through the encryption key, and the encryption key is unique.
6. The Yishitong information processing and analysis system according to claim 5, characterized in that: The step of randomly generating a corresponding encryption key based on the target letter, the first number, and the second number includes: randomly permuting and combining the target letter, the first number, and the second number to generate a number of original sequence passwords with preset digits, and screening the number of original sequence passwords to generate a number of corresponding target sequence passwords; and randomly selecting one of the target sequence passwords as the encryption key.
7. The Yishitong information processing and analysis system according to claim 6, characterized in that: The step of screening the plurality of original sequence codes comprises: detecting one by one the first number corresponding to the numbers and the second number corresponding to the letters in each of the original sequence codes, and determining in real time whether the first number is less than the second number; If it is determined in real time that the first number is smaller than the second number, the original sequence code corresponding to the current first number being smaller than the second number is set as the target sequence code.
8. The Yishitong information processing and analysis system according to claim 1, characterized in that: The classification analysis is specifically as follows: Extracting semantic features of patient identification information; constructing a loss function based on the semantic features, and continuously optimizing the loss function to obtain a text information classification model; The table architecture is classified and parsed using a text information classification model in combination with patient identification information.
9. The Yishitong information processing and analysis system according to claim 8, characterized in that: The semantic features of patient identification information are extracted as follows: Use the bag-of-words model to convert patient identification information text into word vectors; The word vectors are sequentially input into the convolutional layer and the pooling layer to extract the structural features of the table architecture; Segment the patient identification information text and calculate the position weight based on the position of the word in the table structure; Calculate the TF-IDF value of each word; Based on the Gaussian distribution principle, the TF-IDF value and position weight are integrated to calculate the comprehensive weight of the word and obtain the local features; The structural features are fused with the local features to obtain the semantic features of the patient identification information; Construct a loss function based on semantic features and obtain a table architecture classification model after optimization; Use the classification model to classify the table schema and generate data indexes.