An intelligent management system for electronic medical record big data

By calculating the character probability offset and classification in electronic medical record data, and using the same state table for compression, the problem of time-consuming construction of state tables in traditional technology is solved, and the data compression rate and efficiency are improved.

CN118762792BActive Publication Date: 2024-12-06济南科汛智能科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411237708.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-06
Estimated Expiration
2044-09-05

AI Technical Summary

Technical Problem

Traditional finite state entropy encoding takes a long time to build a state table, and the large number of packets leads to an increase in the number of state tables, affecting the efficiency of data compression.

Method used

By calculating the character probability offset in each group, the data is classified, so that the character probability in each class of clusters is close, the same state table is used for compression, and the optimal state is calculated through the probability distribution in each class of clusters. table, reduce the number of state tables and build time.

Benefits of technology

Improves data compression rate, reduces the time to build state tables, and optimizes data storage and transmission efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118762792B_ABST
    Figure CN118762792B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of data processing, and in particular to an intelligent management system for electronic medical record big data, the system comprising: a data acquisition module, a data grouping module, a probability offset calculation module, a grouping classification module, and a data compression storage module, wherein the probability offset of each character in each electronic medical record data subsequence is obtained; the probability offset of each electronic medical record data subsequence is obtained according to the probability offset of each type of character in each electronic medical record data subsequence and the data is classified; multiple types of data are obtained by classification, and the probability corresponding to each type of character in the optimal state table is obtained according to the probability of each type of character in all electronic medical record data subsequences in each type of data, and the electronic medical record data is compressed according to the optimal state table to obtain compressed data, and the compressed data is stored. By using the probability offset to classify the data, the characters in each type of data are distributed more evenly, thereby achieving the purpose of efficient compression.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to an intelligent management system for electronic medical record big data. Background Art

[0002] With the development of medical informatization, electronic medical records have become an important tool for medical institution management and doctor diagnosis. However, with the widespread application of electronic medical record data, the amount of case data continues to increase, which brings huge challenges to data storage and management. In order to solve this problem, data compression technology has been introduced into the electronic medical record big data intelligent management system. By using data compression technology, storage space can be optimized, data transmission efficiency can be improved, data security can be enhanced, and data availability can be improved, thereby achieving efficient management and utilization of electronic medical record data.

[0003] Traditional finite state entropy coding compresses data by constructing a state table based on the frequency of occurrence of data characters. For data with too large a volume, the data in the state table can be reduced by grouping the data. However, the more groups there are, the more state tables need to be constructed, which takes a lot of time to construct.

[0004] The present invention calculates the probability offset of each group according to the probability distribution of each character in the group, classifies the groups according to the probability offset, makes the probability of each character in each cluster approximately equal, and then calculates the optimal state table of each cluster according to the character probability in each cluster. Summary of the invention

[0005] In order to solve the above problems, the present invention provides an electronic medical record big data intelligent management system, the system comprising:

[0006] The data collection module acquires electronic medical record data, which includes the patient's personal information, medical history, diagnosis results, treatment plan, medication status, surgical records, and test results;

[0007] The data grouping module groups the electronic medical record data into several subsequences and counts the probability of each type of character in each subsequence of the electronic medical record data;

[0008] A probability offset calculation module, which obtains the probability offset of each character in each electronic medical record data subsequence according to the difference between the probability of each character in each electronic medical record data subsequence and the mean probability of each character in all electronic medical record data subsequences;

[0009] A grouping and classification module obtains a probability offset of each electronic medical record data subsequence according to a probability offset of each type of character in each electronic medical record data subsequence;

[0010] Acquire the first category of data and the second category of data according to the electronic medical record data subsequence probability offset of each electronic medical record data subsequence;

[0011] Obtain the probability mean corresponding to each character on each electronic medical record data subsequence in all electronic medical record data subsequences, and obtain the classification probability offset of each electronic medical record data subsequence in the second category of data according to the average value of the difference between the probability offset of each category of characters in each electronic medical record data subsequence and the probability mean; divide the electronic medical record data subsequence in the second category of data into a plurality of subcategories of data according to the classification probability offset;

[0012] The data compression storage module obtains the probability corresponding to each type of character in the optimal state table according to the probability of each type of character in all electronic medical record data subsequences in each type of data, compresses the electronic medical record data according to the optimal state table, obtains compressed data, and stores the compressed data.

[0013] Preferably, the step of obtaining the probability offset of each electronic medical record data subsequence according to the probability offset of each type of characters in each electronic medical record data subsequence comprises the following steps:

[0014] The probability offset of a character in each electronic medical record data subsequence is calculated as:

[0015]

[0016] In the formula, Indicates The first The probability offset of the class character, Indicates The first The probability corresponding to the class character; Indicates the total number of electronic medical record data subsequences; Indicates the first The mean probability of the class character; Indicates the first The probability of the class character; Indicates the first The probability of the class character is the same as the probability of the first The maximum absolute value of the difference between the mean probabilities of class characters.

[0017] Preferably, the step of obtaining the probability offset of each character in each electronic medical record data subsequence according to the difference between the probability of each character in each electronic medical record data subsequence and the mean probability of each character in all electronic medical record data subsequences comprises the following steps:

[0018] The calculation formula for the probability offset of each electronic medical record data subsequence is as follows:

[0019]

[0020] In the formula, Indicates The probability offset of the electronic medical record data subsequence of the electronic medical record data subsequence; Indicates The first The probability offset of the class character; Indicates The number of character types in an electronic medical record data subsequence.

[0021] Preferably, the step of acquiring the first type of data and the second type of data according to the electronic medical record data subsequence probability offset of each electronic medical record data subsequence comprises the following steps:

[0022] A probability offset threshold of an electronic medical record data subsequence is set, and all electronic medical record data subsequences whose probability offsets are less than or equal to the probability offset threshold of the electronic medical record data subsequence are recorded as first-category data, and the remaining data are recorded as second-category data.

[0023] Preferably, the method of obtaining the probability mean corresponding to each character on each electronic medical record data subsequence in all electronic medical record data subsequences, and obtaining the classification probability offset of each electronic medical record data subsequence in the second category of data according to the average value of the difference between the probability offset of each type of character in each electronic medical record data subsequence and the probability mean, comprises the following steps:

[0024] The specific calculation formula is as follows:

[0025]

[0026] In the formula, Indicates the second type of data The classification probability offset of the electronic medical record data subsequence; Indicates the second type of data The probability of the mth character in the electronic medical record data subsequence; Represents the number of all character types in the electronic medical record data sequence; Indicates The first Class character; Indicates the first The first subsequence of the electronic medical record data The probability corresponding to the class character; Represents the total number of electronic medical record data subsequences.

[0027] Preferably, the step of dividing the electronic medical record data subsequence in the second category of data into a plurality of subcategory data according to the classification probability offset comprises the following steps:

[0028] Set multiple classification probability offset thresholds; divide the electronic medical record data subsequences in the second category of data into multiple categories through multiple classification probability offset thresholds: for the electronic medical record data subsequences in the second category of data, divide the electronic medical record data subsequences whose classification probability offsets belong to the corresponding interval into one subcategory of data, thereby obtaining multiple subcategory data.

[0029] Preferably, the method of obtaining the probability corresponding to each type of character in the optimal state table according to the probability of each type of character in all electronic medical record data subsequences in each type of data includes the following steps:

[0030] The specific process of the probability corresponding to each type of character in the optimal state table is:

[0031]

[0032] In the formula, Indicates the optimal state table The probability corresponding to the class character; Represents the first category of electronic medical record data EMR data subsequences; Indicates the total number of electronic medical record data subsequences in the first category of data; Indicates The first The probability of the class character.

[0033] Preferably, the step of grouping the electronic medical record data to obtain a plurality of subsequences and counting the probability of each type of character in each electronic medical record data subsequence comprises the following steps:

[0034] The collected electronic medical record data sequence is initially grouped into The data is grouped into 1 group. In the last group, if the amount of data in the group is insufficient When the number of remaining is calculated, the initial grouping of the electronic medical record data sequence is completed. The time series data sequence of the energy storage power supply is recorded as Electronic medical record data subsequence; traverse the electronic medical record data sequence, obtain the character types in the electronic medical record data sequence, sort all the character types in the electronic medical record data sequence according to the dictionary sorting method, and count the number of occurrences of each type of characters in each electronic medical record data subsequence. The ratio of the number of occurrences of each type of characters in each electronic medical record data subsequence to the total number of occurrences of all characters in each electronic medical record data subsequence is the probability of each type of character in each electronic medical record data subsequence.

[0035] Preferably, the method for obtaining the plurality of sub-category data is:

[0036] Set the classification probability offset threshold , , respectively, are recorded as the first classification probability offset threshold The offset threshold of the second classification probability , passing the first classification probability offset threshold The offset threshold of the second classification probability The second type of electronic medical record data subsequences are divided into three categories: for the electronic medical record data subsequences in the second type of data, the classification probability offset belongs to the interval (-1, ) is divided into the first sub-class data, and the classification probability offset belongs to the interval ( ) is divided into the second sub-class data, and the classification probability offset belongs to the interval ( ,The electronic medical record data subsequence within 1) is classified as the third subcategory of data.

[0037] Preferably, the step of compressing the electronic medical record data according to the optimal state table to obtain compressed data includes the following steps:

[0038] By using the optimal state table corresponding to each type of electronic medical record data subsequence data, finite state entropy coding is used to encode and convert each type of electronic medical record data subsequence data to obtain compressed data of each type of data.

[0039] The present invention has the following beneficial effects: by analyzing and processing the character probabilities in each group, the distribution of the same character in each group is obtained, the probability offset of each group is calculated according to the distribution of the same character in each group, and the data groups are classified according to the probability offset, so that the probabilities of the corresponding characters in each cluster are close, so that the data in the same cluster can be compressed using the same state table. The data is classified by calculating the probability offset in each group, so that the compression rate of each cluster using the same state table is higher, and then the optimal state table corresponding to the cluster is calculated according to the probability distribution of each character in each cluster, thereby reducing the number of state tables that need to be constructed and reducing the time required to construct the state table. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A system block diagram of an electronic medical record big data intelligent management system provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of an electronic medical record big data intelligent management system proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0043] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0044] The specific scheme of an electronic medical record big data intelligent management system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0045] See also Figure 1 , which shows an electronic medical record big data intelligent management system provided by an embodiment of the present invention, the system includes the following modules:

[0046] The data collection module 101 is used to collect electronic medical record data.

[0047] Electronic medical record data is collected by manual input, scanning of paper medical records, importing data from other systems, etc. The electronic medical record data includes the patient's personal information, medical history, diagnosis results, treatment plans, medication conditions, surgical records, test results and other medical related information, thereby obtaining an electronic medical record data sequence. The electronic medical record data sequence contains a variety of character types, including numbers, Chinese characters, letters, and special symbols such as punctuation marks.

[0048] Since there are a large number of repeated characters in the electronic medical record data sequence, for example, the names of certain drugs in the medication information will be repeated on a large scale, and the same is true for other data. Repeated characters need to be compressed when stored to save storage space. Existing compression algorithms, such as finite state entropy coding, use the probability of characters for compression. Existing finite state entropy coding only relies on the probability of repeated characters when constructing the state table. In this embodiment, the distribution of characters in the electronic medical record data sequence also needs to be considered. Therefore, this embodiment combines the distribution characteristics of the characters in the electronic medical record data sequence to segment and uniformity analyze the electronic medical record data sequence to ensure that a better compression effect can be obtained in the existing compression algorithm.

[0049] The data grouping module 102 is used to obtain electronic medical record data subsequences.

[0050] It should be noted that when constructing a state table, if the amount of data to be compressed is too large, the number of states is too large, and the time consumed in constructing the state table is longer. Therefore, it is necessary to group the data to reduce the number of states in the state table. However, grouping the data will increase the number of state tables. Therefore, it is necessary to classify the groups with similar data probabilities into one category, so that the data of this category can be compressed using the same filling table, thereby reducing the time consumed by frequently constructing the state table for data with similar probabilities. Then, the optimal state table is adaptively calculated through the probability distribution of the data in each cluster, so that the data in the same cluster can be compressed using the same state table, thereby reducing the number of filling tables.

[0051] The collected electronic medical record data sequence is initially grouped. Since the amount of electronic medical record data sequence data is large, the initial grouping is performed according to the preset threshold. In this embodiment, As an example, other values ​​may be set in specific implementation. This embodiment does not specifically limit the grouping. Every 100 data are grouped into one group. In the last group, if the amount of data in the group is less than 100, the remaining amount is calculated based on the amount. In this way, the initial grouping of the electronic medical record data sequence is completed. The time series data sequence of the energy storage power supply is recorded as Electronic medical record data subsequence; traverse the electronic medical record data sequence, obtain the character types in the electronic medical record data sequence, and sort all the character types in the electronic medical record data sequence according to the dictionary sorting method, where the number of all character types is recorded as , and obtain the probability of each type of character in each electronic medical record data subsequence in the order of dictionary sorting. The method for obtaining the probability of each type of character in each electronic medical record data subsequence is: counting the number of occurrences of each type of character in each electronic medical record data subsequence, and the ratio of the number of occurrences of each type of character in each electronic medical record data subsequence to the total number of occurrences of all characters in each electronic medical record data subsequence is the probability of each type of character in each electronic medical record data subsequence.

[0052] The probability offset calculation module 103 is used to obtain the probability offset of each type of character in the electronic medical record data subsequence.

[0053] It should be noted that the same state table can compress multiple data, but the closer the probability of each character in the state table is to the probability of the data to be compressed, the better the compression effect. Therefore, when reducing the use of state tables, it is necessary to consider the probability distribution of each character in the electronic medical record data subsequence to calculate the probability offset of each group of data. The first The probability of the first character in all electronic medical record data subsequences The greater the difference in the average values ​​of the character probabilities, the The larger the probability offset of the electronic medical record data subsequence, the The first The probability of the first character in all electronic medical record data subsequences The smaller the difference in the average values ​​of the character probabilities, the The smaller the probability offset of the electronic medical record data subsequence.

[0054] The probability offset of a character in each electronic medical record data subsequence is obtained according to the difference between the probability of a character in each electronic medical record data subsequence and the mean probability of a character in all electronic medical record data subsequences. The probability offset of a character in each electronic medical record data subsequence is calculated as follows:

[0055]

[0056] In the formula, Indicates The first The probability offset of the class character, Indicates The first The probability corresponding to the class character; Indicates the total number of electronic medical record data subsequences; Indicates the first The mean probability of the class character; Indicates the first The probability of the class character; Indicates the first The probability of the class character is the same as the probability of the first The maximum absolute value of the difference between the mean values ​​of the class character probabilities; the probability offsets of all class characters in all electronic medical record data subsequences are obtained by the above calculation method.

[0057] It should be noted that: Indicates The first The probability of the class character is the same as the probability of the first The absolute value of the difference between the mean probability of the class characters is larger, that is, the larger the offset is, the higher the probability of the first The more discrete the probability distribution of the class character is, the smaller the value is, that is, the smaller the offset is, which means that the first The probability distribution of class characters is denser. Indicates the first The probability of the class character is the same as the probability of the first The maximum absolute value of the difference between the mean probabilities of the class characters. Each class of characters has only one such value. That is to The result of normalization. Therefore, The larger the value of Group The larger the probability offset of the class character, The smaller the value of Group The probability offset of the class character is smaller.

[0058] The grouping and classification module 104 is used to classify the electronic medical record data subsequences into different categories.

[0059] It should be noted that: by calculating the probability offset of each type of character in each electronic medical record data subsequence, the probability offset of each electronic medical record data subsequence can be obtained according to the probability offset of the same type of characters in all electronic medical record data subsequences. When calculating the probability offset of each electronic medical record data subsequence, if the probability offset of each type of character in each electronic medical record data subsequence is smaller, it means that the probability offset of the electronic medical record data subsequence is smaller, that is, the electronic medical record data subsequence has more similar electronic medical record data subsequences; if the probability offset of each type of character in each electronic medical record data subsequence is larger, it means that the probability offset of the electronic medical record data subsequence is larger, that is, the number of electronic medical record data subsequences similar to the electronic medical record data subsequence is smaller.

[0060] The probability offset of each electronic medical record data subsequence is obtained according to the probability offset of each type of character in each electronic medical record data subsequence. The calculation formula of the probability offset of each electronic medical record data subsequence is as follows:

[0061]

[0062] In the formula, Indicates The probability offset of the electronic medical record data subsequence of the electronic medical record data subsequence; Indicates The first The probability offset of the class character; Indicates The number of character types in the electronic medical record data subsequence. Among them, the smaller the probability offset of each sub-character in the electronic medical record data subsequence, that is, The smaller it is, the smaller the probability offset of the electronic medical record data subsequence is; the larger the probability offset of each type of character in the electronic medical record data subsequence is, that is, The larger it is, the greater the probability offset of the electronic medical record data subsequence.

[0063] Each electronic medical record data subsequence is classified according to the electronic medical record data subsequence probability offset of each electronic medical record data subsequence.

[0064] It should be noted that: the smaller the probability offset of a single electronic medical record data subsequence, the smaller the difference between the probability of each character in the electronic medical record data subsequence and the corresponding characters in the other electronic medical record data subsequences, and the better the compression effect when the state table is generated according to its probability, that is, the less memory the compressed data occupies; the larger the probability offset of a single electronic medical record data subsequence, the larger the difference between the probability of each character in the electronic medical record data subsequence and the corresponding characters in the other electronic medical record data subsequences, and the worse the compression effect when the state table is generated according to its probability, that is, the more memory the compressed data occupies. Therefore, it is necessary to classify the electronic medical record data subsequences according to the probability offset of each electronic medical record data subsequence, so that the probability offsets of each character in each electronic medical record data subsequence are relatively close; the closer the probability offsets of each character in each electronic medical record data subsequence are, the better the compression effect is when the electronic medical record data subsequence data in each category is compressed using the same state table.

[0065] The specific method for classifying data according to the size of the probability offset of the electronic medical record data subsequence in the electronic medical record data subsequence is:

[0066] Set the probability offset threshold of the electronic medical record data subsequence, denoted as In this embodiment, As an example, other values ​​may be set in specific implementation. This embodiment does not specifically limit the value. All electronic medical record data subsequence probability offsets are less than or equal to the electronic medical record data subsequence probability offset threshold. The electronic medical record data subsequences are divided into one category, recorded as the first category data, and the remaining data are recorded as the second category data;

[0067] Obtain the probability mean corresponding to each character in each electronic medical record data subsequence in all electronic medical record data subsequences, and obtain the classification probability offset of each electronic medical record data subsequence in the second category of data based on the average value of the difference between the probability offset of each type of character in each electronic medical record data subsequence and the probability mean. The specific calculation formula is as follows:

[0068]

[0069] In the formula, Indicates the second type of data The classification probability offset of the electronic medical record data subsequence; Indicates the second type of data The probability of the mth character in the electronic medical record data subsequence; Represents the number of all character types in the electronic medical record data sequence; Indicates The first Class character; Indicates the first The first subsequence of the electronic medical record data The probability corresponding to the class character; Represents the total number of electronic medical record data subsequences.

[0070] Among them, according to , that is, The classification probability offset of the electronic medical record data subsequences is used to divide the second type of electronic medical record data subsequences into three categories, which are recorded as the first subclass data, the second subclass data and the third subclass data respectively; The smaller it is, the smaller the probability of each character in the subsequence of the electronic medical record data is. The larger the value, the greater the probability of each character in the electronic medical record data subsequence.

[0071] At this point, the classification probability offset of each electronic medical record data subsequence in the second category of data is obtained.

[0072] Set the classification probability offset threshold , , respectively, are recorded as the first classification probability offset threshold The offset threshold of the second classification probability In this embodiment, As an example, other values ​​can be set in specific implementation; through the first classification probability offset threshold The offset threshold of the second classification probability The second category of electronic medical record data subsequences are divided into three categories: for the electronic medical record data subsequences in the second category of data, the electronic medical record data subsequences whose classification probability offsets belong to the interval (-1, -0, 3) are divided into the first subcategory data, and the electronic medical record data subsequences whose classification probability offsets belong to the interval ( ) is divided into the second sub-category data, and the electronic medical record data sub-sequence whose classification probability offset belongs to the interval (0,3,1) is divided into the third sub-category data.

[0073] The data compression storage module 105 is used to compress and store each type of data.

[0074] It should be noted that after classifying the data according to the probability offset, the corresponding probabilities of the electronic medical record data subsequences in each category are made as equal as possible. When constructing the state table, one state table can compress multiple groups of data, thereby reducing the number of required state tables and the time consumed in constructing the state table. Different state tables have different compression effects on the same group of data. The closer the probability of each character in the state table is to the probability of the data to be compressed, the better the compression effect. The greater the difference between the probability of each character in the state table and the probability of the data to be compressed, the worse the compression effect.

[0075] According to the probability of each type of character in all electronic medical record data subsequences in each type, the probability corresponding to each type of character in the optimal state table is obtained. Taking the first type of data as an example, the specific process is as follows:

[0076]

[0077] In the formula, Indicates the optimal state table The probability corresponding to the class character; Represents the first category of electronic medical record data EMR data subsequences; Indicates the total number of electronic medical record data subsequences in the first category of data; Indicates The first Similarly, the corresponding probabilities in the optimal state table of all types of characters in the first category of data are calculated, and the optimal state table of the first category of data is obtained; similarly, the optimal state tables of the first subcategory data, the second subcategory data, and the third subcategory data are obtained.

[0078] It should be noted that: the electronic medical record data subsequences are obtained and the probability offset in each electronic medical record data subsequence is calculated to classify the electronic medical record data subsequences, so that the probabilities of the corresponding characters in each category are close, so as to maximize the compression rate when constructing the optimal state table; and the optimal state table is constructed according to the probability distribution of each character in each category.

[0079] Utilizing the optimal state table corresponding to each category of electronic medical record data subsequence data, finite state entropy coding is used to encode and convert each category of electronic medical record data subsequence data to obtain compressed data of each category of data, wherein finite state entropy coding is a prior art and will not be described in detail in this embodiment. The compressed data is stored in the electronic medical record big data intelligent management system to facilitate subsequent queries on this part of the data.

[0080] In summary, the system of the present invention includes a data acquisition module, a data grouping module, a probability offset calculation module, a grouping classification module, and a data compression storage module. By analyzing and processing the character probabilities in each group, the distribution of the same character in each group is obtained, and the probability offset of each group is calculated according to the distribution of the same character in each group. The data groups are classified according to the probability offset, so that the probabilities of the corresponding characters in each cluster are close, so that the data in the same cluster can be compressed using the same state table. The data is classified by calculating the probability offset in each group, so that the compression rate of each cluster using the same state table is higher, and then the optimal state table corresponding to the cluster is calculated according to the probability distribution of each character in each cluster, thereby reducing the number of state tables that need to be constructed and reducing the time required to construct the state table.

[0081] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent management system for electronic medical record big data, characterized in that: The system comprises: The data collection module acquires electronic medical record data, which includes the patient's personal information, medical history, diagnosis results, treatment plan, medication status, surgical records, and test results; The data grouping module groups the electronic medical record data into several subsequences and counts the probability of each type of character in each subsequence of the electronic medical record data; A probability offset calculation module, which obtains the probability offset of each character in each electronic medical record data subsequence according to the difference between the probability of each character in each electronic medical record data subsequence and the mean probability of each character in all electronic medical record data subsequences; A grouping and classification module obtains a probability offset of each electronic medical record data subsequence according to a probability offset of each type of character in each electronic medical record data subsequence; Acquire the first category of data and the second category of data according to the electronic medical record data subsequence probability offset of each electronic medical record data subsequence; Obtain the probability mean corresponding to each character on each electronic medical record data subsequence in all electronic medical record data subsequences, and obtain the classification probability offset of each electronic medical record data subsequence in the second category of data according to the average value of the difference between the probability offset of each category of characters in each electronic medical record data subsequence and the probability mean; divide the electronic medical record data subsequence in the second category of data into a plurality of subcategories of data according to the classification probability offset; The data compression storage module obtains the probability corresponding to each type of character in the optimal state table according to the probability of each type of character in all electronic medical record data subsequences in each type of data, compresses the electronic medical record data according to the optimal state table, obtains compressed data, and stores the compressed data; The method of obtaining the probability offset of each electronic medical record data subsequence according to the probability offset of each type of characters in each electronic medical record data subsequence comprises the following steps: The probability offset of each character in each electronic medical record data subsequence is calculated as: In the formula, Indicates The first The probability offset of the class character, Indicates The first The probability corresponding to the class character; Indicates the total number of electronic medical record data subsequences; Indicates the first The mean probability of the class character; Indicates the first The probability of the class character; Indicates the first The probability of the class character is the same as the probability of the first The maximum absolute value of the difference between the mean probabilities of class characters; The step of obtaining the probability offset of each character in each electronic medical record data subsequence according to the difference between the probability of each character in each electronic medical record data subsequence and the probability mean of each character in all electronic medical record data subsequences comprises the following steps: The calculation formula for the probability offset of each electronic medical record data subsequence is as follows: In the formula, Indicates The probability offset of the electronic medical record data subsequence of the electronic medical record data subsequence; Indicates The first The probability offset of the class character; Indicates The number of character types in the EMR data subsequence; The step of obtaining the first type of data and the second type of data according to the electronic medical record data subsequence probability offset of each electronic medical record data subsequence comprises: Setting an electronic medical record data subsequence probability offset threshold, marking all electronic medical record data subsequences whose electronic medical record data subsequence probability offset is less than or equal to the electronic medical record data subsequence probability offset threshold as first-category data, and marking the remaining data as second-category data; The method of obtaining the probability mean corresponding to each character on each electronic medical record data subsequence in all electronic medical record data subsequences, and obtaining the classification probability offset of each electronic medical record data subsequence in the second category of data according to the average value of the difference between the probability offset of each type of character in each electronic medical record data subsequence and the probability mean, comprises the following steps: The specific calculation formula is as follows: In the formula, Indicates the second type of data The classification probability offset of the electronic medical record data subsequence; Indicates the second type of data The probability of the mth character in the electronic medical record data subsequence; Represents the number of all character types in the electronic medical record data sequence; Indicates The first Class character; Indicates the first The first subsequence of the electronic medical record data The probability corresponding to the class character; Represents the total number of electronic medical record data subsequences.

2. The electronic medical record big data intelligent management system according to claim 1 is characterized in that: The step of dividing the electronic medical record data subsequence in the second category of data into a plurality of subcategory data according to the classification probability offset comprises the following steps: Set multiple classification probability offset thresholds; divide the electronic medical record data subsequences in the second category of data into multiple categories through multiple classification probability offset thresholds: for the electronic medical record data subsequences in the second category of data, divide the electronic medical record data subsequences whose classification probability offsets belong to the corresponding interval into one subcategory of data, thereby obtaining multiple subcategory data.

3. The electronic medical record big data intelligent management system according to claim 1 is characterized in that: The method of obtaining the probability corresponding to each type of character in the optimal state table according to the probability of each type of character in all electronic medical record data subsequences in each type of data includes the following steps: The specific process of the probability corresponding to each type of character in the optimal state table is: In the formula, Indicates the optimal state table The probability corresponding to the class character; Represents the first category of electronic medical record data EMR data subsequences; Indicates the total number of electronic medical record data subsequences in the first category of data; Indicates The first The probability of the class character.

4. The electronic medical record big data intelligent management system according to claim 1, characterized in that: The electronic medical record data is grouped to obtain a plurality of subsequences, and the probability of each type of character in each electronic medical record data subsequence is counted, including the following steps: The collected electronic medical record data sequence is initially grouped into The data is grouped into 1 group. In the last group, if the amount of data in the group is insufficient When the number of remaining is calculated, the initial grouping of the electronic medical record data sequence is completed. The time series data sequence of the energy storage power supply group is recorded as Electronic medical record data subsequence; traverse the electronic medical record data sequence, obtain the character types in the electronic medical record data sequence, sort all the character types in the electronic medical record data sequence according to the dictionary sorting method, and count the number of occurrences of each type of characters in each electronic medical record data subsequence. The ratio of the number of occurrences of each type of characters in each electronic medical record data subsequence to the total number of occurrences of all characters in each electronic medical record data subsequence is the probability of each type of character in each electronic medical record data subsequence.

5. The electronic medical record big data intelligent management system according to claim 2, characterized in that: The method for obtaining the several sub-category data is as follows: Set the classification probability offset threshold , , respectively, are recorded as the first classification probability offset threshold The offset threshold of the second classification probability , passing the first classification probability offset threshold The offset threshold of the second classification probability The second type of electronic medical record data subsequences are divided into three categories: for the electronic medical record data subsequences in the second type of data, the classification probability offset belongs to the interval (-1, ) is divided into the first sub-class data, and the classification probability offset belongs to the interval ( ) is divided into the second sub-class data, and the classification probability offset belongs to the interval ( ,The electronic medical record data subsequence within 1) is classified as the third subcategory of data.

6. The electronic medical record big data intelligent management system according to claim 1, characterized in that: The step of compressing the electronic medical record data according to the optimal state table to obtain compressed data includes: By using the optimal state table corresponding to each type of electronic medical record data subsequence data, finite state entropy coding is used to encode and convert each type of electronic medical record data subsequence data to obtain compressed data of each type of data.