Medical data intelligent supervision system and method based on big data

Abstract

The invention discloses an intelligent medical data supervision system and method based on big data, and relates to the technical field of data supervision, and the supervision method comprises the following steps: extracting a feature set of each illness state diagnosis record, and classifying each illness state diagnosis record; analyzing the page view and record generation time of a certain illness state diagnosis record, calculating to obtain a characteristic value, and determining a characteristic diagnosis record; randomly selecting one type of diagnosis record set, and extracting and analyzing a feature set of each illness state diagnosis record to obtain a correlation degree between the illness state diagnosis records; when a medical worker inputs a section of illness state information on the illness state data storage platform for searching, feature analysis is performed on the illness state information, and a corresponding diagnosis record set is matched; and analyzing the correlation degree between the illness state information and each illness state diagnosis record in the diagnosis record set, and generating a push priority for pushing the illness state diagnosis records to medical staff.

Description

Technical Field

[0001] The present invention relates to the technical field of data supervision, and in particular to a medical data intelligent supervision system and method based on big data. Background Art

[0002] Medical data refers to data related to life, health and medical care, covering multiple aspects, including but not limited to clinical data, health monitoring data, biological data and operational data; the collection and analysis of this data is of great significance for improving the quality of medical services, optimizing hospital management, promoting medical technology innovation, and conducting public health monitoring.

[0003] When doctors determine treatment plans for patients, they usually review historical medical records, which can help them make more accurate treatments. Especially in complex cases, they need to rely on historical diagnostic data to help doctors provide the most effective treatments. However, as the number of medical records increases, it becomes increasingly difficult for doctors to find targeted plans. Generally, they can only screen for the same disease, but cannot subdivide different conditions. For specific conditions, they can only review them one by one, which seriously affects the doctor's diagnostic process and efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a medical data intelligent supervision system and method based on big data to solve the problems raised in the prior art.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solution: a medical data intelligent supervision method based on big data, the supervision method comprising the following steps:

[0006] Step S100: Establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets;

[0007] Step S200: arbitrarily obtain the number of visits and the record generation time of a certain disease diagnosis record, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set;

[0008] Step S300: randomly selecting a type of diagnosis record set, extracting a feature set of each disease diagnosis record in the diagnosis record set, analyzing the difference between the diagnosis information corresponding to each feature set, and obtaining the correlation degree between each disease diagnosis record;

[0009] Step S400: When medical staff input a piece of medical information on the medical data storage platform for search, the medical information is feature analyzed and matched with the corresponding diagnosis record set; the correlation between the medical information and each medical diagnosis record in the diagnosis record set is analyzed, and a push priority for pushing the medical diagnosis record to the medical staff is generated.

[0010] Furthermore, step S100 includes the following steps:

[0011] Step S101: Establish a medical terminology database to store a number of medical terms; medical terms include words such as "disease name", "symptoms", "medication instructions", etc., but do not contain specific disease names and specific disease information; obtain the diagnosis information of the i-th disease diagnosis record, segment the diagnosis information, and generate a number of words;

[0012] Step S102: randomly selecting a word from the plurality of words, and performing a similarity comparison with the medical terms in the medical term database; setting a similarity threshold, and if the similarity is greater than the similarity threshold, setting the word as a feature of the i-th disease diagnosis record; by comparing the feature with the medical term, the disease diagnosis record can be effectively divided, so as to facilitate the subsequent similarity between different disease diagnosis records;

[0013] Step S103: Filter out all features from the plurality of words to generate a feature set for the i-th disease diagnosis record; match any remaining word in the words except all features with any feature, set a correlation threshold, and if the correlation is greater than the correlation threshold, set the any remaining word as feature information of the any feature, and obtain feature information corresponding to each feature in the feature set; for example, it is possible to confirm the specific disease corresponding to the disease name;

[0014] Step S104: Establish a disease name database, compare the characteristic information of the ith disease diagnosis record with the disease name in the disease name database, and if there is a characteristic information that is identical to the disease name, set the feature corresponding to the characteristic information as the reference feature of the ith disease diagnosis record; select any two disease diagnosis records from the disease data storage platform, and if the characteristic information corresponding to the reference feature in the two disease diagnosis records is the same, classify the two disease diagnosis records into the same category, and generate a category of diagnosis record set.

[0015] Further, step S200 includes the following steps:

[0016] Step S201: Read the i-th disease diagnosis record whose generation time is D i, the time interval between the generation of the i-th diagnosis record and the current time is T i ; Get the number of visits to the i-th disease diagnosis record at the current time as V i , the access frequency of the i-th diagnosis record is calculated to be f i =V i / T i ;

[0017] Step S202: Obtain the diagnosis record set where the i-th disease diagnosis record is located, and obtain the access frequency of each disease diagnosis record in the diagnosis record set; count the number of disease diagnosis records in the diagnosis record set, and obtain the average access frequency F of the diagnosis record set. i ;

[0018] Step S203: Sort the disease diagnosis records in the diagnosis record set from low to high according to the access frequency to generate a first feature sequence; Sort the disease diagnosis records in the diagnosis record set from recent to ancient according to the record generation time to generate a second feature sequence; Obtain the position L of the i-th disease diagnosis record in the first feature sequence i1 and the position L in the second signature sequence i2 ; According to the formula:

[0019]

[0020] Calculate the characteristic value C of the i-th disease diagnosis record i ;

[0021] By comparing the access frequency with the average access frequency of similar disease diagnosis records, the difference between them reflects the bias of a disease diagnosis record in being accessed among similar disease diagnosis records; too low a bias reflects that the disease diagnosis record will not be used as a reference target by medical staff, so the feature value will be lower; the record generation time is also an influencing factor, because the number of visits will accumulate over time; among disease diagnosis records with high similarity, newer diagnosis records are often selected because there may be improvements in diagnostic solutions. Therefore, in addition to analyzing the access frequency, it is also necessary to balance the relationship between the access volume and the record generation time to obtain a more scientific and accurate feature value; which is conducive to the subsequent setting of push priorities;

[0022] Step S204: Obtain the characteristic value of each disease diagnosis record in the diagnosis record set, and select the disease diagnosis record with the largest characteristic value as the characteristic diagnosis record of the diagnosis record set; set the characteristic diagnosis record to be the record most frequently accessed by medical staff among the same type of diagnosis records, indicating that when searching for similar diseases, medical staff will give priority to referring to the characteristic diagnosis record, which is convenient for the subsequent push priority setting.

[0023] Further, step S300 includes the following steps:

[0024] Step S301: arbitrarily obtain the i-th disease diagnosis record and the i'th disease diagnosis record in a certain type of diagnosis record set, and obtain feature sets of the two disease diagnosis records respectively; arbitrarily extract one feature from the two feature sets for comparison, and if the two extracted features are the same, set the extracted feature as the common feature of the two disease diagnosis records, and obtain the common feature set of the two disease diagnosis records;

[0025] Step S302: Select the jth common feature in the common feature set, obtain the feature information corresponding to the jth common feature in the two disease diagnosis records, compare the two feature information, and calculate the similarity of the feature information corresponding to the jth common feature in the two disease diagnosis records.

[0026] Step S303: Obtain the number of feature sets in the i-th disease diagnosis record as M i And the number of feature sets in the i'th disease diagnosis record is M i′ , the number of remaining features in the i-th disease diagnosis record is calculated to be (M i ) remain =M i -M i,i′ And the number of remaining features in the i'th disease diagnosis record is (M i′ ) remain =M i′ -M i,i′ , where M i,i′ is the number of common features in the common feature set;

[0027] Step S304: Obtain the similarity of the feature information corresponding to each common feature in the two disease diagnosis records; according to the formula:

[0028]

[0029] Calculate the correlation degree G between the i-th disease diagnosis record and the i'th disease diagnosis record i,i′; Analyzing the similarities between the same features can reflect the correlation between the two disease diagnosis records, while analyzing the proportions of different features can reflect the differences between the two disease diagnosis records; only by taking both the correlation and the difference into account at the same time can the degree of difference be more accurate.

[0030] Furthermore, step S400 includes the following steps:

[0031] Step S401: Obtain the condition information input by the medical staff and generate a feature set of the condition information; arbitrarily select one feature from the feature set, extract feature information corresponding to the feature, and compare it with the disease name in the disease name database to obtain a reference feature of the condition information, and determine the diagnosis record set corresponding to the condition information on the condition data storage platform;

[0032] Step S402: arbitrarily select the i-th disease diagnosis record from the diagnosis record set, and obtain the correlation degree G between the disease information and the i-th disease diagnosis record. i ; Sort the diagnosis records in the diagnosis record set into queues according to the degree of association from large to small;

[0033] Step S403: extract characteristic diagnosis records from the diagnosis record set, and obtain the degree of correlation between the condition information and the characteristic diagnosis record as G0; set a correlation degree threshold θ, if G0 ≥ θ, move the characteristic diagnosis record as the first condition diagnosis record in the queue; generate the push priority of the condition data storage platform to push the condition diagnosis record to medical staff; when the degree of correlation between the information that the medical staff wants to query and the characteristic diagnosis record is higher, the characteristic diagnosis record is directly selected as the reference target, and only when the correlation degree is not high, it will be pushed according to the correlation degree.

[0034] In order to better implement the above method, a medical data intelligent supervision system based on big data is also proposed. The supervision system includes a historical diagnosis classification module, a feature record analysis module, an information association analysis module and a real-time push analysis module;

[0035] The historical diagnosis classification module is used to establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets;

[0036] The characteristic record analysis module is used to obtain the number of visits and the record generation time of a certain disease diagnosis record at will, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set;

[0037] An information association analysis module is used to select a type of diagnosis record set at random, extract a feature set of each disease diagnosis record in the diagnosis record set, analyze the difference between the diagnosis information corresponding to each feature set, and obtain the degree of association between each disease diagnosis record;

[0038] The real-time push analysis module is used to perform feature analysis on the medical information and match the corresponding diagnosis record set when the medical staff inputs a piece of medical information for search on the medical data storage platform; analyze the degree of correlation between the medical information and each medical diagnosis record in the diagnosis record set, and generate a push priority for pushing the medical diagnosis record to the medical staff.

[0039] Further, the historical diagnosis classification module includes a diagnosis record generation unit and a diagnosis record classification unit;

[0040] The diagnosis record generation unit is used to establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; the diagnosis record classification unit is used to extract the feature set of each disease diagnosis record, classify each disease diagnosis record, and generate several types of diagnosis record sets.

[0041] Further, the feature record analysis module includes a data analysis unit and a feature calculation unit;

[0042] The data analysis unit is used to arbitrarily obtain the number of visits and the record generation time of a certain medical diagnosis record, and analyze the relationship between the number of visits and the record generation time of the certain medical diagnosis record; the feature calculation unit is used to calculate the feature value of the certain medical diagnosis record and determine the feature diagnosis record in each type of diagnosis record set.

[0043] Further, the information association analysis module includes a difference analysis unit and an association calculation unit;

[0044] The difference analysis unit is used to arbitrarily select a type of diagnosis record set, extract the feature set of each disease diagnosis record in the diagnosis record set, and analyze the differences between the diagnosis information corresponding to each feature set; the association calculation unit is used to obtain the degree of association between each disease diagnosis record based on the differences between the diagnosis information corresponding to each feature set.

[0045] Furthermore, the real-time push analysis module includes an information matching unit and a record push unit;

[0046] The information matching unit is used to perform feature analysis on the medical information and match the corresponding diagnosis record set when the medical staff inputs a piece of medical information for search on the medical data storage platform; the record pushing unit is used to analyze the degree of correlation between the medical information and each medical diagnosis record in the diagnosis record set, and generate a push priority for pushing the medical diagnosis record to the medical staff.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. The present invention classifies the disease diagnosis records and evaluates the reference value of each disease diagnosis record according to the access frequency of each disease diagnosis record; when the disease diagnosis records are subsequently pushed to medical staff, the push priority is set according to the degree of association and the reference value, so as to meet the needs of medical staff with the greatest efficiency and improve the diagnosis efficiency of medical staff;

[0049] 2. The present invention calculates the characteristic value of each medical diagnosis record by factors such as the access frequency and record generation time of each medical diagnosis record, reflecting that the medical diagnosis record with a higher characteristic value has a more reference value, which is convenient for subsequent setting of push priority and helps medical staff find the required reference records as quickly as possible;

[0050] 3. The present invention helps medical staff match the most similar diagnosis records by comparing the correlation between disease diagnosis records; improves the diagnosis efficiency of medical staff and helps medical staff provide more accurate diagnosis methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of the steps of an intelligent supervision method for medical data based on big data;

[0052] Figure 2 This is a structural diagram of an intelligent medical data supervision system based on big data. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0054] Example: Figure 1 to Figure 2As shown, the present invention provides a medical data intelligent supervision method based on big data, and the supervision method comprises the following steps:

[0055] Step S100: Establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets;

[0056] Wherein, step S100 includes the following steps:

[0057] Step S101: Establish a medical terminology database to store a number of medical terms; obtain the diagnosis information of the i-th disease diagnosis record, segment the diagnosis information, and generate a number of words;

[0058] Step S102: randomly selecting a word from the plurality of words and comparing the similarity with the medical terms in the medical term database; setting a similarity threshold, and if the similarity is greater than the similarity threshold, setting the word as a feature of the i-th disease diagnosis record;

[0059] Step S103: Filter out all features from the plurality of words to generate a feature set for the i-th disease diagnosis record; match any remaining word in the words except all features with any feature in terms of correlation, set a correlation threshold, and if the correlation is greater than the correlation threshold, set the any remaining word as feature information of the any feature, and obtain feature information corresponding to each feature in the feature set;

[0060] Step S104: Establish a disease name database, compare the characteristic information of the ith disease diagnosis record with the disease name in the disease name database, and if there is a characteristic information that is identical to the disease name, set the feature corresponding to the characteristic information as the reference feature of the ith disease diagnosis record; select any two disease diagnosis records from the disease data storage platform, and if the characteristic information corresponding to the reference feature in the two disease diagnosis records is the same, classify the two disease diagnosis records into the same category, and generate a category of diagnosis record set.

[0061] Embodiment 1: When two disease diagnosis records are matched with the disease names in the disease name database and both are the same as a certain disease, it means that the two disease diagnosis records are both treatment plans for the certain disease, and therefore the two disease diagnosis records are classified into the same category.

[0062] Step S200: arbitrarily obtain the number of visits and the record generation time of a certain disease diagnosis record, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set;

[0063] Wherein, step S200 includes the following steps:

[0064] Step S201: Read the i-th disease diagnosis record whose generation time is D i , the time interval between the generation of the i-th diagnosis record and the current time is T i ; Get the number of visits to the i-th disease diagnosis record at the current time as V i , the access frequency of the i-th diagnosis record is calculated to be f i =V i / T i ;

[0065] Step S202: Obtain the diagnosis record set where the i-th disease diagnosis record is located, and obtain the access frequency of each disease diagnosis record in the diagnosis record set; count the number of disease diagnosis records in the diagnosis record set, and obtain the average access frequency F of the diagnosis record set. i ;

[0066] Step S203: Sort the disease diagnosis records in the diagnosis record set from low to high according to the access frequency to generate a first feature sequence; Sort the disease diagnosis records in the diagnosis record set from recent to ancient according to the record generation time to generate a second feature sequence; Obtain the position L of the i-th disease diagnosis record in the first feature sequence i1 and the position L in the second signature sequence i2 ; According to the formula:

[0067]

[0068] Calculate the characteristic value C of the i-th disease diagnosis record i ;

[0069] Example 2: Set the number of accesses to record A to 100 times, and the time interval between the generation time of record A and the current time to 100 days; sort other disease diagnosis records of the same type as record A, set the position in the first feature sequence to 6, the position in the second feature sequence to 2, and the average access frequency to 0.5; therefore, the characteristic value of record A is calculated to be 100×6 / 2×(1+0.5 / 0.5)=600;

[0070] Step S204: obtaining a characteristic value of each disease diagnosis record in the diagnosis record set, and selecting a disease diagnosis record with the largest characteristic value as a characteristic diagnosis record of the diagnosis record set.

[0071] Step S300: randomly selecting a type of diagnosis record set, extracting a feature set of each disease diagnosis record in the diagnosis record set, analyzing the difference between the diagnosis information corresponding to each feature set, and obtaining the correlation degree between each disease diagnosis record;

[0072] Wherein, step S300 includes the following steps:

[0073] Step S301: arbitrarily obtain the i-th disease diagnosis record and the i'th disease diagnosis record in a certain type of diagnosis record set, and obtain feature sets of the two disease diagnosis records respectively; arbitrarily extract one feature from the two feature sets for comparison, and if the two extracted features are the same, set the extracted feature as the common feature of the two disease diagnosis records, and obtain the common feature set of the two disease diagnosis records;

[0074] Step S302: Select the jth common feature in the common feature set, obtain the feature information corresponding to the jth common feature in the two disease diagnosis records, compare the two feature information, and calculate the similarity of the feature information corresponding to the jth common feature in the two disease diagnosis records.

[0075] Step S303: Obtain the number of feature sets in the i-th disease diagnosis record as M i And the number of feature sets in the i'th disease diagnosis record is M i′ , the number of remaining features in the i-th disease diagnosis record is calculated to be (M i ) remain =M i -M i,i′ And the number of remaining features in the i'th disease diagnosis record is (M i′ ) remain =M i′ -M i,i′ , where M i,i′ is the number of common features in the common feature set;

[0076] Step S304: Obtain the similarity of the feature information corresponding to each common feature in the two disease diagnosis records; according to the formula:

[0077]

[0078] Calculate the correlation degree G between the i-th disease diagnosis record and the i'th disease diagnosis recordi,i′ ;

[0079] Example 3: Set the number of features of the feature set of record A to 5, the number of features of the feature set of feature B to 6, and the number of common features of the two records to 3; set the similarities of the three common features to 100%, 90%, and 80%, respectively; calculate the degree of association between record A and record B to be (1+0.9+0.8) / 3×(1-5 / 11)=0.9×(1-0.455)=0.49;

[0080] Step S400: When a medical staff inputs a piece of medical information for search on the medical data storage platform, the medical information is feature analyzed and matched with a corresponding diagnosis record set; the correlation between the medical information and each medical diagnosis record in the diagnosis record set is analyzed, and a push priority for pushing the medical diagnosis record to the medical staff is generated;

[0081] Wherein, step S400 includes the following steps:

[0082] Step S401: Obtain the condition information input by the medical staff and generate a feature set of the condition information; arbitrarily select one feature from the feature set, extract feature information corresponding to the feature, and compare it with the disease name in the disease name database to obtain a reference feature of the condition information, and determine the diagnosis record set corresponding to the condition information on the condition data storage platform;

[0083] Step S402: arbitrarily select the i-th disease diagnosis record from the diagnosis record set, and obtain the correlation degree G between the disease information and the i-th disease diagnosis record. i ; Sort the diagnosis records in the diagnosis record set into queues according to the degree of association from large to small;

[0084] Step S403: extracting characteristic diagnosis records from the diagnosis record set, obtaining the correlation degree between the condition information and the characteristic diagnosis record as G0; setting a correlation degree threshold θ, if G0 ≥ θ, moving the characteristic diagnosis record as the first condition diagnosis record in the queue; generating a push priority for the condition data storage platform to push the condition diagnosis record to medical staff;

[0085] Example 4: Set the correlation degree threshold to 80%. If the correlation degree between the characteristic diagnosis record and the condition information input by the medical staff is greater than 80%, the characteristic diagnosis record will be pushed first, and the remaining condition diagnosis records will be pushed in order; if the correlation degree is less than 80%, they will be directly sorted from large to small according to the correlation degree and pushed.

[0086] An intelligent medical data supervision system based on big data, the supervision system includes a historical diagnosis classification module, a feature record analysis module, an information association analysis module and a real-time push analysis module;

[0087] The historical diagnosis classification module is used to establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets;

[0088] The characteristic record analysis module is used to obtain the number of visits and the record generation time of a certain disease diagnosis record at will, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set;

[0089] An information association analysis module is used to select a type of diagnosis record set at random, extract a feature set of each disease diagnosis record in the diagnosis record set, analyze the difference between the diagnosis information corresponding to each feature set, and obtain the degree of association between each disease diagnosis record;

[0090] The real-time push analysis module is used to perform feature analysis on the medical information and match the corresponding diagnosis record set when the medical staff inputs a piece of medical information for search on the medical data storage platform; analyze the degree of correlation between the medical information and each medical diagnosis record in the diagnosis record set, and generate a push priority for pushing the medical diagnosis record to the medical staff.

[0091] Among them, the historical diagnosis classification module includes a diagnosis record generation unit and a diagnosis record classification unit;

[0092] The diagnosis record generation unit is used to establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; the diagnosis record classification unit is used to extract the feature set of each disease diagnosis record, classify each disease diagnosis record, and generate several types of diagnosis record sets.

[0093] Among them, the feature record analysis module includes a data analysis unit and a feature calculation unit;

[0094] The data analysis unit is used to arbitrarily obtain the number of visits and the record generation time of a certain medical diagnosis record, and analyze the relationship between the number of visits and the record generation time of the certain medical diagnosis record; the feature calculation unit is used to calculate the feature value of the certain medical diagnosis record and determine the feature diagnosis record in each type of diagnosis record set.

[0095] Wherein, the information association analysis module includes a difference analysis unit and an association calculation unit;

[0096] The difference analysis unit is used to arbitrarily select a type of diagnosis record set, extract the feature set of each disease diagnosis record in the diagnosis record set, and analyze the differences between the diagnosis information corresponding to each feature set; the association calculation unit is used to obtain the degree of association between each disease diagnosis record based on the differences between the diagnosis information corresponding to each feature set.

[0097] Among them, the real-time push analysis module includes an information matching unit and a record push unit;

[0098] The information matching unit is used to perform feature analysis on the medical information and match the corresponding diagnosis record set when the medical staff inputs a piece of medical information for search on the medical data storage platform; the record pushing unit is used to analyze the degree of correlation between the medical information and each medical diagnosis record in the diagnosis record set, and generate a push priority for pushing the medical diagnosis record to the medical staff.

[0099] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for intelligent supervision of medical data based on big data, characterized by: The supervision method comprises the following steps: Step S100: Establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets; Step S200: arbitrarily obtain the number of visits and the record generation time of a certain disease diagnosis record, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set; Step S300: randomly selecting a type of diagnosis record set, extracting a feature set of each disease diagnosis record in the diagnosis record set, analyzing the difference between the diagnosis information corresponding to each feature set, and obtaining the correlation degree between each disease diagnosis record; Step S400: When medical staff input a piece of medical information on the medical data storage platform for search, the medical information is feature analyzed and matched with the corresponding diagnosis record set; the correlation between the medical information and each medical diagnosis record in the diagnosis record set is analyzed, and a push priority for pushing the medical diagnosis record to the medical staff is generated.

2. The intelligent medical data supervision method based on big data according to claim 1 is characterized by: The step S100 includes the following steps: Step S101: Establish a medical terminology database to store a number of medical terms; obtain the diagnosis information of the i-th disease diagnosis record, segment the diagnosis information, and generate a number of words; Step S102: randomly selecting a word from the plurality of words and comparing the similarity with the medical terms in the medical term database; setting a similarity threshold, and if the similarity is greater than the similarity threshold, setting the word as a feature of the i-th disease diagnosis record; Step S103: Filter out all features from the plurality of words to generate a feature set for the i-th disease diagnosis record; match any remaining word in the words except all features with any feature in terms of correlation, set a correlation threshold, and if the correlation is greater than the correlation threshold, set the any remaining word as feature information of the any feature, and obtain feature information corresponding to each feature in the feature set; Step S104: Establish a disease name database, compare the characteristic information of the ith disease diagnosis record with the disease name in the disease name database, and if there is a characteristic information that is identical to the disease name, set the feature corresponding to the characteristic information as the reference feature of the ith disease diagnosis record; select any two disease diagnosis records from the disease data storage platform, and if the characteristic information corresponding to the reference feature in the two disease diagnosis records is the same, classify the two disease diagnosis records into the same category, and generate a category of diagnosis record set.

3. The intelligent medical data supervision method based on big data according to claim 2 is characterized by: The step S200 includes the following steps: Step S201: Read the i-th disease diagnosis record whose generation time is D i , the time interval between the generation of the i-th diagnosis record and the current time is T i ; Get the number of visits to the i-th disease diagnosis record at the current time as V i , the access frequency of the i-th diagnosis record is calculated to be f i =V i / T i ; Step S202: Obtain the diagnosis record set where the i-th disease diagnosis record is located, and obtain the access frequency of each disease diagnosis record in the diagnosis record set; count the number of disease diagnosis records in the diagnosis record set, and obtain the average access frequency F of the diagnosis record set. i ; Step S203: Sort the disease diagnosis records in the diagnosis record set from low to high according to the access frequency to generate a first feature sequence; Sort the disease diagnosis records in the diagnosis record set from recent to ancient according to the record generation time to generate a second feature sequence; Obtain the position L of the i-th disease diagnosis record in the first feature sequence i1 and the position L in the second signature sequence i2 ; According to the formula: Calculate the characteristic value C of the i-th disease diagnosis record i ; Step S204: obtaining a characteristic value of each disease diagnosis record in the diagnosis record set, and selecting a disease diagnosis record with the largest characteristic value as a characteristic diagnosis record of the diagnosis record set.

4. The intelligent medical data supervision method based on big data according to claim 3 is characterized by: The step S300 includes the following steps: Step S301: arbitrarily obtain the i-th disease diagnosis record and the i'th disease diagnosis record in a certain type of diagnosis record set, and obtain feature sets of the two disease diagnosis records respectively; arbitrarily extract one feature from the two feature sets for comparison, and if the two extracted features are the same, set the extracted feature as the common feature of the two disease diagnosis records, and obtain the common feature set of the two disease diagnosis records; Step S302: Select the jth common feature in the common feature set, obtain the feature information corresponding to the jth common feature in the two disease diagnosis records, compare the two feature information, and calculate the similarity of the feature information corresponding to the jth common feature in the two disease diagnosis records. Step S303: Obtain the number of feature sets in the i-th disease diagnosis record as M i And the number of feature sets in the i'th disease diagnosis record is M i′ , the number of remaining features in the i-th disease diagnosis record is calculated to be (M i ) remain =M i -M i,i′ And the number of remaining features in the i'th disease diagnosis record is (M i′ ) remain =M i′ -M i,i′ , where M i,i′ is the number of common features in the common feature set; Step S304: Obtain the similarity of the feature information corresponding to each common feature in the two disease diagnosis records; according to the formula: Calculate the correlation degree G between the i-th disease diagnosis record and the i'th disease diagnosis record i,i′ .

5. The intelligent medical data supervision method based on big data according to claim 4 is characterized by: The step S400 includes the following steps: Step S401: Obtain the condition information input by the medical staff and generate a feature set of the condition information; arbitrarily select one feature from the feature set, extract feature information corresponding to the feature, and compare it with the disease name in the disease name database to obtain a reference feature of the condition information, and determine the diagnosis record set corresponding to the condition information on the condition data storage platform; Step S402: arbitrarily select the i-th disease diagnosis record from the diagnosis record set, and obtain the correlation degree G between the disease information and the i-th disease diagnosis record. i ; Sort the diagnosis records in the diagnosis record set into queues according to the degree of association from large to small; Step S403: extract characteristic diagnosis records from the diagnosis record set, and obtain the degree of association between the condition information and the characteristic diagnosis record as G0; set a correlation degree threshold θ, if G0≥θ, move the characteristic diagnosis record as the first condition diagnosis record in the queue; generate the push priority of the condition data storage platform to push the condition diagnosis record to medical staff.

6. A medical data intelligent supervision system, used to execute a medical data intelligent supervision method based on big data as claimed in any one of claims 1 to 5, characterized in that: The supervision system includes a historical diagnosis classification module, a feature record analysis module, an information association analysis module and a real-time push analysis module; The historical diagnosis classification module is used to establish a disease data storage platform, record each disease diagnosis process of all patients, and generate corresponding disease diagnosis records; extract the feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets; The characteristic record analysis module is used to obtain the number of visits and the record generation time of a certain disease diagnosis record at random, analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record, obtain the characteristic value of the certain disease diagnosis record, and determine the characteristic diagnosis record in each type of diagnosis record set; The information association analysis module is used to arbitrarily select a type of diagnosis record set, extract the feature set of each disease diagnosis record in the diagnosis record set, analyze the difference between the diagnosis information corresponding to each feature set, and obtain the degree of association between each disease diagnosis record; The real-time push analysis module is used to perform feature analysis on a piece of medical information and match it with a corresponding diagnosis record set when the medical staff inputs a piece of medical information on the medical data storage platform for search; analyze the degree of correlation between the medical information and each medical diagnosis record in the diagnosis record set, and generate a push priority for pushing medical diagnosis records to the medical staff.

7. The intelligent medical data supervision system according to claim 6, characterized in that: The historical diagnosis classification module includes a diagnosis record generation unit and a diagnosis record classification unit; The diagnosis record generation unit is used to establish a disease data storage platform to record each disease diagnosis process of all patients and generate corresponding disease diagnosis records; the diagnosis record classification unit is used to extract a feature set of each disease diagnosis record, and classify each disease diagnosis record to generate several types of diagnosis record sets.

8. The intelligent medical data supervision system according to claim 6, characterized in that: The feature record analysis module includes a data analysis unit and a feature calculation unit; The data analysis unit is used to obtain the number of visits and the record generation time of a certain disease diagnosis record at random, and analyze the relationship between the number of visits and the record generation time of the certain disease diagnosis record; The feature calculation unit is used to calculate the feature value of the certain disease diagnosis record and determine the feature diagnosis record in each type of diagnosis record set.

9. The intelligent medical data supervision system according to claim 6, characterized in that: The information association analysis module includes a difference analysis unit and an association calculation unit; The difference analysis unit is used to arbitrarily select a type of diagnostic record set, extract the feature set of each disease diagnosis record in the diagnostic record set, and analyze the differences between the diagnostic information corresponding to each feature set; the association calculation unit is used to obtain the degree of association between each disease diagnosis record based on the differences between the diagnostic information corresponding to each feature set.

10. The intelligent medical data supervision system according to claim 6, characterized in that: The real-time push analysis module includes an information matching unit and a record push unit; The information matching unit is used to perform feature analysis on a piece of medical information and match it with a corresponding set of diagnostic records when the medical staff inputs the medical information for search on the medical data storage platform; the record pushing unit is used to analyze the degree of association between the medical information and each medical diagnosis record in the diagnostic record set, and generate a push priority for pushing the medical diagnosis record to the medical staff.

Images