CN112133429B - Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium - Google Patents

Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium Download PDF

Info

Publication number
CN112133429B
CN112133429B CN202011030257.3A CN202011030257A CN112133429B CN 112133429 B CN112133429 B CN 112133429B CN 202011030257 A CN202011030257 A CN 202011030257A CN 112133429 B CN112133429 B CN 112133429B
Authority
CN
China
Prior art keywords
diagnosis
user
treatment
review
nth
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202011030257.3A
Other languages
Chinese (zh)
Other versions
CN112133429A (en
Inventor
向正贵
张伟平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Taikang Health Industry Investment Holdings Co ltd
Taikang Insurance Group Co Ltd
Original Assignee
Taikang Health Industry Investment Holdings Co ltd
Taikang Insurance Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Taikang Health Industry Investment Holdings Co ltd, Taikang Insurance Group Co Ltd filed Critical Taikang Health Industry Investment Holdings Co ltd
Priority to CN202011030257.3A priority Critical patent/CN112133429B/en
Publication of CN112133429A publication Critical patent/CN112133429A/en
Application granted granted Critical
Publication of CN112133429B publication Critical patent/CN112133429B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
    • G06Q30/0224—Discounts or incentives, e.g. coupons or rebates based on user history
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

The invention discloses a diagnosis and treatment prediction method and device, computer equipment and a computer readable storage medium, wherein the method comprises the following steps: firstly, acquiring consumption medical data of a user, and further determining diagnosis and treatment period portrait of the user according to the consumption medical data of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point; and finally, determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item. According to the invention, the diagnosis and treatment period portrait of the user is determined through the consumption medical data, and then the Nth re-diagnosis time interval of each diagnosis and treatment item of the user is determined through data fitting and maximum likelihood estimation, so that the accuracy of re-diagnosis interval prediction can be improved.

Description

Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a diagnosis and treatment prediction method and apparatus, a computer device, and a computer readable storage medium.
Background
This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims. The description herein is not admitted to be prior art by inclusion in this section.
In some cases, doctors will require or suggest that the patient visit the hospital for review at a timely time based on their own professional judgment. The diagnosis and treatment scheme is more dependent on the professional experience of doctors, and meanwhile, the doctors determine the re-diagnosis matters according to various factors such as illness states. For example, the diagnosis project and the diagnosis timing of the review, the predicted cost of the review, and the like.
Besides the mode of judging the re-diagnosis and treatment time by the professional doctor, patients can be regarded as consumers (medical institution clients) at present, the client value is determined by an RFM (Recency, frequency, monetary) model, and further the re-diagnosis time and re-diagnosis project of the patients are determined according to the value of the patients. In the clinical medical field, the RFM model divides the value of a client into eight quadrant dimensions according to three factors, namely the latest time of visit (return), the Frequency of visit (Frequency) and the amount of visit (monetariy) of the client, as described in the following table.
Time of recent visit (R) Frequency of treatment (F) Treatment amount (M) Customer category
Is now close to Multiple ones Big size High value customer
Far from now Multiple ones Big size Important protection of clients
Is now close to Less quantity Big size Key development client
Far from now Less quantity Big size The key point is to save the customer
Is now close to Multiple ones Small size General value customer
Far from now Multiple ones Small size Typically keep customers
Is now close to Less quantity Small size General development clients
Far from now Less quantity Small size Potential customers
Table-guest value RFM model
As can be seen from the above table one, the customers with the most frequent visits and the large amount of visits are high-value customers, and the customers with the most frequent visits and the small amount of visits are general-value customers.
It follows that the RFM model has the advantages of simplicity, intuitiveness and convenience in calculation, but the RFM model also has obvious disadvantages. The RFM model is fuzzy in three-dimensional eight-quadrant division of the client value, does not accurately and quantitatively analyze factors such as the change of the client value along with time, can judge the diagnosis and treatment project or approximate diagnosis and treatment time of the re-diagnosis only according to the client value in the RFM model, and cannot accurately determine the diagnosis and treatment information of the re-diagnosis.
Therefore, the conventional method for determining the diagnosis and treatment information of the patient by using the RFM model has the problem of poor accuracy.
Disclosure of Invention
The embodiment of the invention provides a diagnosis and treatment prediction method, which is used for improving accuracy of the prediction of a re-diagnosis interval, and comprises the following steps:
acquiring consumption medical data of a user;
determining a diagnosis and treatment period portrait of the user according to the consumption medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point;
determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item;
wherein M is a natural number, N is a positive integer, and M is less than N.
The embodiment of the invention also provides a diagnosis and treatment prediction device, which is used for improving accuracy of the review interval prediction, and comprises the following steps:
the data acquisition module is used for acquiring the consumption medical data of the user;
the portrait determining module is used for determining the portrait of the diagnosis and treatment period of the user according to the consumption medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point;
The interval determining module is used for determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item;
wherein M is a natural number, N is a positive integer, and M is less than N.
The embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the diagnosis and treatment prediction method is realized when the processor executes the computer program.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program for executing the diagnosis and treatment prediction method.
In the embodiment of the invention, firstly, the consumption medical data of a user is obtained, and then the diagnosis and treatment period portrait of the user is determined according to the consumption medical data of the user; and finally, determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item. According to the embodiment of the invention, the diagnosis and treatment period portrait of the user is determined through consuming the medical data, and then the Nth review time interval of each diagnosis and treatment item of the user is determined through data fitting and maximum likelihood estimation, so that the accuracy of review interval prediction can be improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
FIG. 1 is a flowchart of an implementation of a diagnosis and treatment prediction method provided by an embodiment of the present invention;
fig. 2 is a flowchart illustrating implementation of step 102 in the diagnosis and treatment prediction method according to an embodiment of the present invention;
fig. 3 is a flowchart illustrating implementation of step 103 in the diagnosis and treatment prediction method according to the embodiment of the present invention;
fig. 4 is a flowchart illustrating implementation of step 303 in the diagnosis and treatment prediction method according to the embodiment of the present invention;
FIG. 5 is a flowchart of another implementation of the diagnosis and treatment prediction method according to the embodiment of the present invention;
fig. 6 is a flowchart illustrating implementation of step 501 in the diagnosis and treatment prediction method according to an embodiment of the present invention;
fig. 7 is a flowchart illustrating implementation of step 603 in the diagnosis and treatment prediction method according to an embodiment of the present invention;
FIG. 8 is a functional block diagram of a diagnosis and treatment prediction apparatus according to an embodiment of the present invention;
Fig. 9 is a block diagram illustrating a structure of an image determining module 802 in the diagnosis and treatment prediction apparatus according to an embodiment of the present invention;
fig. 10 is a block diagram illustrating a structure of an interval determining module 803 in the diagnosis and treatment prediction apparatus according to an embodiment of the present invention;
fig. 11 is a block diagram showing the structure of an interval determining unit 1003 in the diagnosis and treatment prediction apparatus according to the embodiment of the present invention;
FIG. 12 is another functional block diagram of a diagnosis and treatment prediction apparatus according to an embodiment of the present invention;
fig. 13 is a block diagram illustrating a construction of a cost determining module 1201 in a diagnosis and treatment predicting apparatus according to an embodiment of the present invention;
fig. 14 is a block diagram showing the structure of the fee determining unit 1303 in the diagnosis and treatment predicting apparatus according to the embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present invention and their descriptions herein are for the purpose of explaining the present invention, but are not to be construed as limiting the invention.
Fig. 1 shows a flow of implementation of the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portions relevant to the embodiment of the present invention are shown, which are described in detail below:
as shown in fig. 1, the diagnosis and treatment prediction method includes:
Step 101, obtaining consumer medical data of a user;
step 102, determining a diagnosis and treatment period portrait of the user according to the consumption medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point;
step 103, determining the nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item.
Wherein M is a natural number, N is a positive integer, and M is less than N.
In an embodiment of the present invention, the consumer medical data of the user may include a diagnosis and treatment record of the user. Upon acquiring consumer medical data of a user, a diagnosis and treatment record of the user may be acquired from at least one medical system. For example, a diagnosis and treatment record of a user is obtained from the same medical system or a plurality of different medical systems, so as to form consumption medical data of the user.
In an embodiment of the present invention, in order to further improve accuracy of the review interval prediction, the diagnosis and treatment prediction method further includes:
Preprocessing the consumer medical data of the user to obtain preprocessed consumer medical data of the user.
The preprocessing may include desensitization processing of customer sensitive information, data processing of null or missing values, conversion processing of special characters, data duplication removal processing, and the like, and those skilled in the art will understand that preprocessing may also include other preprocessing measures besides the above preprocessing measures, such as data merging, data digitizing or discretizing, and the like, and medical diagnosis text parsing, classifying and structuring processing, and the like, which are not particularly limited by the embodiment of the present invention. In addition, the consumer medical data of the user may also include customer base information, customer level, customer insurance category, customer member attributes, customer diagnosis and treatment characteristics, customer diagnosis and treatment profile, and the like.
After the consumption medical data of the user is obtained, a diagnosis and treatment portrait of the user is generated according to various information in the consumption medical data of the user. The diagnosis and treatment portrait of the user at least comprises each diagnosis and treatment item of the user and the previous M times of diagnosis and treatment time intervals of each diagnosis and treatment item. For example, the consumption medical data of the user is used for acquiring each diagnosis and treatment time point of each diagnosis and treatment item of the user, such as a time node of initial diagnosis and a time point of review of each diagnosis and treatment item of the user, and then the previous M times of review time intervals of each diagnosis and treatment item of the user are determined according to the time point of initial diagnosis and the time point of review of each diagnosis and treatment item of the user. The time interval of the re-diagnosis may be a time interval between a time point of the last diagnosis and a time point of the previous diagnosis, or may also be a time interval between a current time point of the diagnosis and a time point of the first diagnosis.
Wherein M is a natural number, i.e. M comprises 0 and positive integers such as 1, 2 and 3. When M is 0, it indicates that the user only makes a first diagnosis and does not make a second diagnosis yet, so that there is no second diagnosis time interval. When M is a positive integer, the user is not only initially diagnosed, but also at least one review, and the diagnosis and treatment period portrait of the user comprises the previous M review time intervals of the diagnosis and treatment item.
In addition, the consumer medical data of the user may include one diagnosis and treatment item, and may further include a plurality of diagnosis and treatment items. The diagnosis and treatment item refers to a diagnosis and treatment item determined according to the conventional definition of the medical field, and generally refers to a diagnosis and treatment item meeting the following conditions: (1) Clinical diagnosis and treatment items which are necessary, safe and effective and have proper cost; (2) Diagnosis and treatment projects with charging standards are formulated by the price department.
In order to accurately predict the re-diagnosis time interval of each diagnosis and treatment item, it is also necessary to acquire the historical re-diagnosis time interval of the historical user of each diagnosis and treatment item. For example, the initial diagnosis time point and the review time point of the historical user of each diagnosis and treatment item are obtained through the consumption medical data of the historical user, and then the historical review time interval of the historical user of each diagnosis and treatment item is determined based on the initial diagnosis time point and the review time point of the historical user of each diagnosis and treatment item. The historical review time interval comprises a review time interval of historical times of diagnosis and treatment records.
After the previous M times of review time intervals of each diagnosis and treatment item of the user and the historical review time intervals of the historical users of each diagnosis and treatment item are determined, the Nth review time interval of each diagnosis and treatment item of the user is determined by utilizing a data fitting and maximum likelihood estimating method, and then the predicted Nth review time interval is utilized to timely inform the user to conduct review in the Nth review time interval.
In the embodiment of the invention, firstly, the consumption medical data of a user is obtained, and then the diagnosis and treatment period portrait of the user is determined according to the consumption medical data of the user; and finally, determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item. According to the embodiment of the invention, the diagnosis and treatment period portrait of the user is determined through consuming the medical data, and then the Nth review time interval of each diagnosis and treatment item of the user is determined through data fitting and maximum likelihood estimation, so that the accuracy of review interval prediction can be improved.
Fig. 2 shows a flow of implementation of step 102 in the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portions relevant to the embodiment of the present invention are shown, which are described in detail below:
In one embodiment of the present invention, to improve the efficiency of determining the medical cycle representation, as shown in fig. 2, step 102, determining the medical cycle representation of the user based on the consumer medical data of the user, includes:
step 201, extracting keywords in the user consumption medical data, and determining each diagnosis and treatment item of the user according to the extracted keywords in the user consumption medical data;
step 202, extracting the diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data, and determining the previous M times of re-diagnosis time intervals of each diagnosis and treatment item according to the extracted diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data;
and 203, determining the diagnosis and treatment period portrait of the user according to each diagnosis and treatment item of the user and the previous M times of re-diagnosis time intervals of each diagnosis and treatment item.
After the consumption medical data of the user is obtained, keywords are extracted from the consumption medical data, and then each diagnosis and treatment item of the user is determined according to the extracted keywords. Specifically, after the keywords in the medical data consumed by the user are extracted, diagnosis and treatment items corresponding to the keywords are determined from the keyword diagnosis and treatment item data table. The keyword diagnosis and treatment item data table comprises keywords, diagnosis and treatment items and corresponding relations between the keywords and the diagnosis and treatment items. The keywords corresponding to the diagnosis and treatment items in the keyword diagnosis and treatment item data table may include one or more keywords. For example, the extracted keywords are matched with keywords in the keyword diagnosis and treatment item data table, keywords in the successfully matched keyword diagnosis and treatment item data table are determined, and then corresponding diagnosis and treatment items are determined according to the keywords in the successfully matched keyword diagnosis and treatment item data table.
In addition, the diagnosis and treatment time of each diagnosis and treatment item of the user is further extracted from the consumption medical data of the user, for example, the initial diagnosis time and the re-diagnosis time of the user in each diagnosis and treatment item, and the first M re-diagnosis time intervals of the user in each diagnosis and treatment item are further determined based on the initial diagnosis time and the re-diagnosis time of the user in each diagnosis and treatment item.
After each diagnosis and treatment item of the user and the previous M times of review time intervals of the user in each diagnosis and treatment item are determined, generating and determining diagnosis and treatment period portraits of the user based on each diagnosis and treatment item of the user and the previous M times of review time intervals of the user in each diagnosis and treatment item.
In the embodiment of the invention, keywords in the user consumption medical data are extracted, each diagnosis and treatment item of the user is determined according to the extracted keywords in the user consumption medical data, the diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data is extracted, the previous M times of re-diagnosis time intervals of each diagnosis and treatment item are determined according to the extracted diagnosis and treatment time of each diagnosis and treatment item of the user, and the diagnosis and treatment period portrait of the user is determined according to each diagnosis and treatment item of the user and the previous M times of re-diagnosis time intervals of each diagnosis and treatment item. According to the embodiment of the invention, each diagnosis and treatment item of the user is determined based on the extracted keywords, the previous M times of re-diagnosis time intervals of each diagnosis and treatment item are determined based on the extracted diagnosis and treatment time, and finally, the diagnosis and treatment period portrait is determined according to each diagnosis and treatment item of the user and the previous M times of re-diagnosis time intervals of each diagnosis and treatment item, so that the efficiency of determining the diagnosis and treatment period portrait can be improved.
Fig. 3 shows a flow of implementing step 103 in the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portions relevant to the embodiment of the present invention are shown, which are described in detail below:
in an embodiment of the present invention, in order to further improve accuracy of review interval prediction, as shown in fig. 3, step 103, determining, according to the previous M review time intervals of each diagnosis and treatment item of the user and the historical review time intervals of the historical user of each diagnosis and treatment item, the nth review time interval of each diagnosis and treatment item of the user by using the maximum likelihood estimation includes:
step 301, according to the previous M times of review time intervals of each diagnosis and treatment item of the user, determining an nth time of review time interval set of each diagnosis and treatment item conforming to normal distribution through data fitting; the nth review time interval set of each diagnosis item comprises an nth review time interval of a historical user of each diagnosis item;
step 302, determining an nth review time interval mean value and an nth review time interval variance of an nth review time interval set of each diagnosis and treatment item by using maximum likelihood estimation;
step 303, determining the nth re-diagnosis time interval of each diagnosis and treatment item of the user according to the nth re-diagnosis time interval mean and the nth re-diagnosis time interval variance.
Specifically, in determining the nth review time interval of each diagnosis and treatment item of the user, based on and according to the previous M review time intervals of each diagnosis and treatment item of the user, obtaining consumption medical data (diagnosis and treatment records) of all users of each diagnosis and treatment item from a database, determining initial review time points and calendar review time points of all users of each diagnosis and treatment item through the diagnosis and treatment records of all users of each diagnosis and treatment item, and further determining the calendar review time intervals of all users of each diagnosis and treatment item.
For each treatment item, the nth set of review time intervals contains two parts: (1) When the historical re-diagnosis time interval of the user comprises the nth re-diagnosis time interval, taking the nth re-diagnosis time interval of the user as the nth re-diagnosis time interval of the user in the diagnosis and treatment item; (2) When the time interval of the last re-diagnosis of the user does not contain the time interval of the nth re-diagnosis, the time interval of the nth re-diagnosis of the user obtained by fitting is obtained by taking the time point of the last re-diagnosis of the user on the diagnosis and treatment item (the initial diagnosis time point and the re-diagnosis time point) as the basis, and the time interval of the nth re-diagnosis of the user on the diagnosis and treatment item is obtained by fitting through a data fitting mode.
In addition, since the nth re-diagnosis time interval of the user on the diagnosis item is not the nth re-diagnosis time interval obtained by fitting, and is not the nth re-diagnosis time interval of the user on the diagnosis item, in order to maintain the data accuracy, the accuracy of the re-diagnosis interval prediction is further improved, and in one embodiment of the present invention, the nth re-diagnosis time interval set only includes the nth re-diagnosis time interval of the user on the diagnosis item in the (1) section, but does not include the nth re-diagnosis time interval of the user on the diagnosis item obtained by fitting in the (2) section. That is, the nth review time interval set only includes the nth review time interval of the first part of users, wherein the historical review time interval of the first part of users is required to include the nth review time interval, and the nth review time interval of the second part of users on the diagnosis and treatment item is empty.
Based on the rule of statistics, the nth review time interval set of each diagnosis and treatment item obtained here, wherein the nth review time interval of the user is an nth review time interval set conforming to the rule of normal distribution.
After the nth re-diagnosis time interval set of each diagnosis and treatment item is determined, determining the nth re-diagnosis time interval mean value and the nth re-diagnosis time interval variance of the nth re-diagnosis time interval set of each diagnosis and treatment item by using a maximum likelihood estimation method.
Further, the nth review time interval of each diagnosis item of the user is determined based on the nth review time interval mean and the nth review time interval variance of the nth review time interval set of each diagnosis item.
In the embodiment of the invention, the nth re-diagnosis time interval set of each diagnosis item conforming to normal distribution is determined through data fitting, the nth re-diagnosis time interval mean value and the nth re-diagnosis time interval variance of the nth re-diagnosis time interval set of each diagnosis item are further determined through maximum likelihood estimation, and finally the nth re-diagnosis time interval of each diagnosis item of a user is determined based on the nth re-diagnosis time interval mean value and the nth re-diagnosis time interval variance, so that the accuracy of re-diagnosis interval prediction can be further improved.
Fig. 4 shows a flow of implementation of step 303 in the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portion relevant to the embodiment of the present invention is shown, which is described in detail below:
In an embodiment of the present invention, in order to improve the flexibility of determining the nth review time interval, as shown in fig. 4, step 303, determining the nth review time interval of each diagnosis item of the user according to the nth review time interval mean and the nth review time interval variance includes:
step 401, determining a target review time interval according to the nth review time interval mean value and the nth review time interval variance;
step 402, when the probability that the nth review time interval of each diagnosis item of the user falls into the target review time interval is not smaller than the preset probability value, taking the target review time interval as the nth review time interval of each diagnosis item of the user.
Let us assume mu respectively 1 Sum sigma 1 Representing the mean value of the nth re-diagnosis time interval and the variance of the nth re-diagnosis time interval, and determining the target re-diagnosis time interval by the following method:
[μ 1 -k 1 σ 1 ,μ 1 +k 1 σ 1 ]；
and satisfy k 1 ≥0；
k 1 Representing interval variance coefficient by adjusting interval variance coefficient k 1 Flexibly adjust the target review time interval.
Wherein, the probability that the nth review time interval of each diagnosis and treatment item of the user falls into the target review time interval is assumed to be p 1 Then at probability p 1 When the probability value is not smaller than the preset probability value, the probability that the nth re-diagnosis time interval of each diagnosis and treatment item of the user falls into the target re-diagnosis time interval is indicated to be quite possible, and the target re-diagnosis time interval can be used as the nth re-diagnosis time interval of each diagnosis and treatment item of the user.
Wherein, at interval, the variance coefficient k 1 Under the known condition, the N-th review time interval of each diagnosis and treatment item of the user can be determined to fall into the target review time interval [ mu ] based on the statistical rule of normal distribution 1 -k 1 σ 1 ,μ 1 +k 1 σ 1 ]Probability p of (2) 1 . The nth review time interval of each diagnosis and treatment item of the user falls into the target review time interval [ mu ] 1 -k 1 σ 1 ,μ 1 +k 1 σ 1 ]Probability p of (2) 1 In the known case, the interval variance coefficient k can be determined based on the statistical regularity of the normal distribution 1 . I.e. interval variance coefficient k 1 Probability p 1 There is a correspondence based on a normal distribution rule.
In addition, a system based on normal distributionThe calculation and probability rules can be known, and the interval variance coefficient k 1 Is an integer greater than 0, such as 1, 2, 3, etc., based on the mean value mu of the nth review interval 1 Variance sigma of time interval of nth re-diagnosis 1 The determined target review time interval is [ mu ] 1 -σ 1 ,μ 1 +σ 1 ]、[μ 1 -2σ 1 ,μ 1 +2σ 1 ]And [ mu ] 1 -3σ 1 ,μ 1 +3σ 1 ]And so on, the nth review time interval of each diagnosis and treatment item of the user falls into the target review time interval [ mu ] 1 -σ 1 ,μ 1 +σ 1 ]、[μ 1 -2σ 1 ,μ 1 +2σ 1 ]And [ mu ] 1 -3σ 1 ,μ 1 +3σ 1 ]The probabilities of (a) are 0.683, 0.955, 0.997, etc., respectively.
The preset probability value is a preset probability value, and it can be understood by those skilled in the art that the preset probability value can be preset according to the actual profile and specific requirements. For example, in particular in one embodiment, [ mu ] will 1 -σ 1 ,μ 1 +σ 1 ]As the target review time interval, the N-th review time interval of each diagnosis and treatment item of the user falls into the target review time interval [ mu ] 1 -σ 1 ,μ 1 +σ 1 ]Is 0.683 as a preset probability value, or will [ mu ] 1 -2σ 1 ,μ 1 +2σ 1 ]As the target review time interval, the N-th review time interval of each diagnosis and treatment item of the user falls into the target review time interval [ mu ] 1 -2σ 1 ,μ 1 +2σ 1 ]Is used as a preset probability value, or [ mu ] is used as a probability of 0.955 1 -3σ 1 ,μ 1 +3σ 1 ]As the target review time interval, the N-th review time interval of each diagnosis and treatment item of the user falls into the target review time interval [ mu ] 1 -3σ 1 ,μ 1 +3σ 1 ]As a preset probability value, etc.
In the embodiment of the invention, the target review time interval is determined according to the average value of the nth review time interval and the variance of the nth review time interval, and when the probability that the nth review time interval of each diagnosis and treatment item of the user falls into the target review time interval is not smaller than the preset probability value, the target review time interval is used as the nth review time interval of each diagnosis and treatment item of the user, and the flexibility of determining the nth review time interval can be improved by flexibly setting the preset probability value.
Fig. 5 shows another implementation flow of the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portion relevant to the embodiment of the present invention is shown, which is described in detail below:
In one embodiment of the invention, the user's medical cycle representation further includes the first M review costs for each medical item of the user. In order to improve accuracy of review cost prediction, as shown in fig. 5, the diagnosis and treatment prediction method further includes, based on the steps of the method:
step 501, according to the first M review costs of each diagnosis and treatment item of the user and the historical review costs of the historical user of each diagnosis and treatment item, determining the nth review cost of each diagnosis and treatment item of the user by using data fitting and maximum likelihood estimation.
In the embodiment of the invention, the diagnosis and treatment period portrait of the user also comprises the first M times of review cost of each diagnosis and treatment item of the user, so that the diagnosis and treatment prediction method can realize the prediction of the Nth review cost of the user on each diagnosis and treatment item.
Specifically, similar to the prediction of the nth review time interval of the user on each diagnosis and treatment item, the calendar diagnosis and treatment cost of each diagnosis and treatment item of the user is firstly extracted from the acquired consumption medical data of the user, and the calendar diagnosis and treatment cost of each diagnosis and treatment item of the user comprises the first M review costs of each diagnosis and treatment item of the user. Further, a medical period portrait of the user on each medical item is generated according to each medical item of the user, the previous M times of review time intervals of each medical item of the user, and the previous M times of review cost of each medical item of the user.
Accordingly, in one embodiment of the present invention, in order to enhance the richness of the medical cycle representation, step 102, determining the medical cycle representation of the user based on the consumer medical data of the user further includes:
extracting the first M times of re-diagnosis cost of the user in each diagnosis and treatment item in the medical data consumed by the user;
step 203, determining a diagnosis and treatment period portrait of the user according to each diagnosis and treatment item of the user and the previous M times of re-diagnosis time intervals of each diagnosis and treatment item, and replacing the diagnosis and treatment period portrait with:
and determining the diagnosis and treatment period portrait of the user according to each diagnosis and treatment item of the user, the previous M times of review time intervals of each diagnosis and treatment item and the previous M times of review cost of each diagnosis and treatment item.
In order to accurately predict the review cost of each diagnosis and treatment item, it is also necessary to obtain the historical review cost of the historical user of each diagnosis and treatment item. For example, the historical user initial diagnosis cost and the historical review cost of each diagnosis and treatment item are obtained through the consumption medical data of the historical user. Further, when predicting the nth review cost of each diagnosis and treat item of the user, determining the nth review cost of each diagnosis and treat item of the user by using data fitting and maximum likelihood estimation according to the previous M review costs of each diagnosis and treat item of the user and the historical review cost of the historical user of each diagnosis and treat item.
In the embodiment of the invention, according to the previous M times of review cost of each diagnosis and treatment item of the user and the historical review cost of the historical user of each diagnosis and treatment item, the Nth time of review cost of each diagnosis and treatment item of the user is determined by utilizing data fitting and maximum likelihood estimation, so that the accuracy of review cost prediction can be improved.
Fig. 6 shows a flow of implementation of step 501 in the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portion relevant to the embodiment of the present invention is shown, which is described in detail below:
in an embodiment of the present invention, in order to improve accuracy of review cost prediction, as shown in fig. 6, step 501, determining, according to the previous M review costs of each diagnosis and treatment item of the user and the historical review costs of the historical user of each diagnosis and treatment item, the nth review cost of each diagnosis and treatment item of the user by using the maximum likelihood estimation includes:
step 601, determining an nth review cost set of each diagnosis item according to normal distribution through data fitting according to the previous M review costs of each diagnosis item of a user; the nth review cost set of each treatment item includes an nth review cost of the historical user of each treatment item;
step 602, determining an nth review cost mean value and an nth review cost variance of an nth review cost set of each diagnosis and treatment item by using the maximum likelihood estimation;
And step 603, determining the nth re-diagnosis cost of each diagnosis and treatment item of the user according to the nth re-diagnosis cost mean value and the nth re-diagnosis cost variance.
Specifically, on the basis of and according to the nth review cost of each diagnosis and treatment item of the user, obtaining consumption medical data (diagnosis and treatment records) of all historical users of each diagnosis and treatment item from a database, and determining the historical review cost of all users of each diagnosis and treatment item according to the diagnosis and treatment records of all historical users of each diagnosis and treatment item.
For each treatment item, the nth review cost set contains two parts: (1) When the historical re-diagnosis cost of the user comprises the nth re-diagnosis cost, taking the nth re-diagnosis cost of the user as the nth re-diagnosis cost of the user in the diagnosis and treatment item; (2) When the historical re-diagnosis cost of the user does not contain the nth re-diagnosis cost, the nth re-diagnosis cost of the user on the diagnosis and treatment item is obtained through fitting by using a data fitting mode based on the historical diagnosis and treatment cost of the user on the diagnosis and treatment item, and the nth re-diagnosis cost of the user obtained through fitting is used as the nth re-diagnosis cost of the user on the diagnosis and treatment item.
In addition, since the nth review cost of the user who does not include the nth review cost in the past is the nth review cost obtained by fitting the nth review cost in the diagnosis and treatment item, and is not the nth review cost of the user who actually exists in the diagnosis and treatment item, in order to maintain the accuracy of the data, the accuracy of the review cost prediction is further improved. That is, the nth review cost set only includes the nth review cost of the first part of users, wherein the historical review cost of the first part of users is required to include the nth review cost, and the nth review cost of the second part of users on the diagnosis and treatment item is empty.
Based on the rule of statistics, the nth review cost set of each diagnosis and treatment item obtained here contains the nth review cost of the user, which is an nth review cost set conforming to the normal distribution rule.
After the nth review cost set of each diagnosis and treatment item is determined, the nth review cost mean value and the nth review cost variance of the nth review cost set of each diagnosis and treatment item are determined by using a maximum likelihood estimation method. Further, the nth review cost of each diagnosis and treatment item of the user is determined based on the nth review cost mean value and the nth review cost variance of the nth review cost set of each diagnosis and treatment item.
According to the embodiment of the invention, according to the previous M times of re-diagnosis cost of each diagnosis and treatment item of a user, the nth time of re-diagnosis cost set of each diagnosis and treatment item which accords with normal distribution is determined through data fitting, the nth time of re-diagnosis cost mean value and the nth time of re-diagnosis cost variance of the nth time of each diagnosis and treatment item set are further determined through maximum likelihood estimation, and finally the nth time of re-diagnosis cost of each diagnosis and treatment item of the user is determined according to the nth time of re-diagnosis cost mean value and the nth time of re-diagnosis cost variance, so that the accuracy of re-diagnosis cost prediction can be improved.
Fig. 7 shows a flow of implementation of step 603 in the diagnosis and treatment prediction method provided by the embodiment of the present invention, and for convenience of description, only the portion relevant to the embodiment of the present invention is shown, which is described in detail below:
in an embodiment of the present invention, in order to improve the flexibility of review cost prediction, as shown in fig. 7, step 603, determining the nth review cost of each diagnosis and treatment item of the user according to the nth review cost mean and the nth review cost variance includes:
step 701, determining a target review cost interval according to the nth review cost mean value and the nth review cost variance;
step 702, when the probability that the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval is not smaller than the preset probability value, taking the target review cost interval as the nth review cost of each diagnosis and treatment item of the user.
Let us assume mu respectively 2 Sum sigma 2 Representing the average value of the nth re-diagnosis cost and the variance of the nth re-diagnosis cost, and determining a target re-diagnosis cost interval by the following method:
[μ 2 -k 2 σ 2 ,μ 2 +k 2 σ 2 ]；
and satisfy k 2 ≥0；
k 2 Representing the cost variance factor by adjusting the cost variance factor k 2 Flexibly adjust the target review cost interval.
Wherein, the probability that the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval is assumed to be p 2 Then at probability p 2 When the probability value is not smaller than the preset probability value, the probability that the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval is indicated to be quite possible, and the target review cost interval can be used as the nth review cost of each diagnosis and treatment item of the user.
Wherein, in the cost variance coefficient k 2 Under the known condition, the N-th review cost of each diagnosis and treatment item of the user can be determined to fall into the target review cost interval [ mu ] based on the statistical rule of normal distribution 2 -k 2 σ 2 ,μ 2 +k 2 σ 2 ]Probability p of (2) 2 . The nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval [ mu ] 2 -k 2 σ 2 ,μ 2 +k 2 σ 2 ]Probability p of (2) 2 In the known case, the cost variance factor k can be determined based on the statistical law of normal distribution 2 . I.e. cost variance factor k 2 Probability p 2 There is a correspondence based on a normal distribution rule.
In addition, based on statistics of normal distribution and probability law, it can be known that in the cost variance coefficient k 1 Is an integer greater than 0, e.g. 1, 2Average value mu of the cost based on the nth review 1 Variance sigma of nth review cost 1 The determined target review cost interval is [ mu ] 2 -σ 2 ,μ 2 +σ 2 ]、[μ 2 -2σ 2 ,μ 2 +2σ 2 ]And [ mu ] 2 -3σ 2 ,μ 2 +3σ 2 ]And so on, the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval [ mu ] 2 -σ 2 ,μ 2 +σ 2 ]、[μ 2 -2σ 2 ,μ 2 +2σ 2 ]And [ mu ] 2 -3σ 2 ,μ 2 +3σ 2 ]The probabilities of (a) are 0.683, 0.955, 0.997, etc., respectively.
The preset probability value is a preset probability value, and it can be understood by those skilled in the art that the preset probability value can be preset according to the actual profile and specific requirements. For example, in particular in one embodiment, [ mu ] will 2 -σ 2 ,μ 2 +σ 2 ]As a target review cost interval, the N-th review cost of each diagnosis and treatment item of the user is fallen into the target review cost interval [ mu ] 2 -σ 2 ,μ 2 +σ 2 ]Is 0.683 as a preset probability value, or will [ mu ] 2 -2σ 2 ,μ 2 +2σ 2 ]As a target review cost interval, the N-th review cost of each diagnosis and treatment item of the user is fallen into the target review cost interval [ mu ] 2 -2σ 2 ,μ 2 +2σ 2 ]Is used as a preset probability value, or [ mu ] is used as a probability of 0.955 2 -3σ 2 ,μ 2 +3σ 2 ]As a target review cost interval, the N-th review cost of each diagnosis and treatment item of the user is fallen into the target review cost interval [ mu ] 2 -3σ 2 ,μ 2 +3σ 2 ]As a preset probability value, etc.
In the embodiment of the invention, the target review cost interval is determined according to the nth review cost mean value and the nth review cost variance, and when the probability that the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval is not smaller than the preset probability value, the target review cost interval is used as the nth review cost of each diagnosis and treatment item of the user, so that the flexibility of review cost prediction can be improved.
In addition, after the user performs the nth re-diagnosis on a certain diagnosis and treatment item, the nth re-diagnosis performed by the user on the diagnosis and treatment item is called a history diagnosis and treatment record, and the accuracy of the prediction of the re-diagnosis time interval is further improved by using the information of the nth re-diagnosis actually occurring on the diagnosis and treatment item, such as the nth re-diagnosis time interval actually occurring by the user, and optimizing the nth re-diagnosis time interval set by using the cross entropy as an optimization objective function; in addition, the cost of the nth review which is actually generated is used for optimizing the cost set of the nth review by taking cross entropy as an optimization objective function, so that the accuracy of review cost prediction is further improved.
In an embodiment of the present invention, in order to implement prediction of total diagnosis and treatment cost of each diagnosis and treatment item of the user, the diagnosis and treatment period portrait of the user further includes a history diagnosis and treatment cost of each diagnosis and treatment item of the user. On the basis of the method steps, the diagnosis and treatment prediction method further comprises the following steps:
and determining the diagnosis and treatment total cost of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation according to the calendar diagnosis and treatment cost of each diagnosis and treatment item of the user and the historical diagnosis and treatment total cost of the historical user of each diagnosis and treatment item.
In an embodiment of the present invention, to improve accuracy of diagnosis and treatment total cost prediction, determining the diagnosis and treatment total cost of each diagnosis and treatment item of a user by using data fitting and maximum likelihood estimation according to the calendar diagnosis and treatment cost of each diagnosis and treatment item of the user and the historical diagnosis and treatment total cost of the historical user of each diagnosis and treatment item, includes:
According to the historical diagnosis and treatment cost of each diagnosis and treatment item of the user, determining a diagnosis and treatment total cost set of each diagnosis and treatment item conforming to normal distribution through data fitting; the diagnosis and treatment total cost set of each diagnosis and treatment item comprises the diagnosis and treatment total cost of the historical user of each diagnosis and treatment item;
determining a diagnosis and treatment total cost mean value and a diagnosis and treatment total cost variance of a diagnosis and treatment total cost set of each diagnosis and treatment item by using the maximum likelihood estimation;
and determining the total diagnosis cost of each diagnosis item of the user according to the average value of the total diagnosis cost and the variance of the total diagnosis cost.
In an embodiment of the present invention, to improve flexibility of total cost prediction, determining total cost of diagnosis for each diagnosis item of a user according to a total cost average and a total cost variance of diagnosis includes:
determining a target diagnosis and treatment total cost interval according to the diagnosis and treatment total cost mean value and the diagnosis and treatment total cost variance;
when the probability that the diagnosis and treatment total cost of each diagnosis and treatment item of the user falls into the target diagnosis and treatment total cost interval is not smaller than the preset probability value, the target diagnosis and treatment total cost interval is used as the diagnosis and treatment total cost of each diagnosis and treatment item of the user.
The prediction of the total cost of diagnosis and treatment of each diagnosis and treatment item of the user is similar to the prediction of the nth diagnosis and treatment cost of each diagnosis and treatment item of the user, and the prediction of the total cost of diagnosis and treatment of each diagnosis and treatment item of the user can be specifically referred to the prediction of the nth diagnosis and treatment cost of each diagnosis and treatment item of the user, and will not be described in detail here.
The embodiment of the invention also provides a diagnosis and treatment prediction device, which is described in the following embodiment. Because the principle of solving the problems by the devices is similar to that of diagnosing and predicting the method, the implementation of the devices can be referred to the implementation of the method, and the repetition is omitted.
Fig. 8 shows functional modules of the diagnosis and treatment prediction apparatus provided by the embodiment of the present invention, and for convenience of explanation, only the portions relevant to the embodiment of the present invention are shown in detail as follows:
referring to fig. 8, each module included in the diagnosis and treatment prediction apparatus is configured to perform each step in the corresponding embodiment of fig. 1, and detailed descriptions of fig. 1 and the corresponding embodiment of fig. 1 will be omitted herein. In the embodiment of the invention, the diagnosis and treatment prediction device comprises a data acquisition module 801, an image determination module 802 and an interval determination module 803.
The data acquisition module 801 is configured to acquire consumer medical data of a user.
A representation determination module 802 for determining a representation of a medical treatment cycle of a user based on consumer medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point.
The interval determining module 803 is configured to determine, according to the previous M review time intervals of each diagnosis and treatment item of the user and the historical review time intervals of the historical user of each diagnosis and treatment item, an nth review time interval of each diagnosis and treatment item of the user by using data fitting and maximum likelihood estimation.
Wherein M is a natural number, N is a positive integer, and M is less than N.
In the embodiment of the invention, firstly, a data acquisition module 801 acquires the consumption medical data of a user, and further a portrait determination module 802 determines a diagnosis and treatment period portrait of the user according to the consumption medical data of the user; and finally, the interval determining module 803 determines the nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item. According to the embodiment of the invention, the diagnosis and treatment period portrait of the user is determined through consuming the medical data, and then the Nth review time interval of each diagnosis and treatment item of the user is determined through data fitting and maximum likelihood estimation, so that the accuracy of review interval prediction can be improved.
In an embodiment of the present invention, in order to further improve accuracy of the review interval prediction, the diagnosis and treatment prediction apparatus further includes: the preprocessing module is used for preprocessing the consumption medical data of the user to obtain preprocessed consumption medical data of the user.
Fig. 9 shows a schematic structural diagram of an image determining module 802 in the diagnosis and treatment prediction apparatus according to the embodiment of the present invention, and for convenience of explanation, only the portions related to the embodiment of the present invention are shown, which is described in detail below:
in an embodiment of the present invention, in order to improve the efficiency of determining the medical period image, referring to fig. 9, each unit included in the image determining module 802 is configured to perform each step in the corresponding embodiment of fig. 2, and detailed descriptions in fig. 2 and the corresponding embodiment of fig. 2 are omitted herein. In the embodiment of the present invention, the portrait determining module 802 includes a keyword extracting unit 901, a diagnosis and treatment time extracting unit 902, and a portrait determining unit 903.
The keyword extraction unit 901 is configured to extract keywords in the medical data consumed by the user, and determine each diagnosis and treatment item of the user according to the extracted keywords in the medical data consumed by the user.
The diagnosis time extracting unit 902 is configured to extract a diagnosis time of each diagnosis item in the user consumption medical data, and determine a first M review time interval of each diagnosis item according to the extracted diagnosis time of each diagnosis item in the user consumption medical data.
And a portrait determining unit 903 for determining a portrait of the diagnosis and treatment period of the user according to each diagnosis and treatment item of the user and the previous M times of review time intervals of each diagnosis and treatment item.
In the embodiment of the present invention, the keyword extraction unit 901 extracts keywords in the user consumption medical data, determines each diagnosis and treatment item of the user according to the extracted keywords in the user consumption medical data, the diagnosis and treatment time extraction unit 902 extracts diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data, determines the first M times of review time intervals of each diagnosis and treatment item according to the diagnosis and treatment time of each diagnosis and treatment item in the extracted user consumption medical data, and the representation determination unit 903 determines a diagnosis and treatment period representation of the user according to each diagnosis and treatment item of the user and the first M times of review time intervals of each diagnosis and treatment item. According to the embodiment of the invention, each diagnosis and treatment item of the user is determined based on the extracted keywords, the previous M times of re-diagnosis time intervals of each diagnosis and treatment item are determined based on the extracted diagnosis and treatment time, and finally, the diagnosis and treatment period portrait is determined according to each diagnosis and treatment item of the user and the previous M times of re-diagnosis time intervals of each diagnosis and treatment item, so that the efficiency of determining the diagnosis and treatment period portrait can be improved.
Fig. 10 shows a schematic structure of an interval determining module 803 in the diagnosis and treatment prediction apparatus according to the embodiment of the present invention, and for convenience of explanation, only the portions related to the embodiment of the present invention are shown, which is described in detail below:
In an embodiment of the present invention, in order to further improve accuracy of the review interval prediction, referring to fig. 10, each unit included in the interval determination module 803 is configured to perform each step in the corresponding embodiment of fig. 3, and detailed descriptions of the corresponding embodiment of fig. 3 are omitted herein. In the embodiment of the present invention, the interval determining module 803 includes a first set determining unit 1001, a first estimation determining unit 1002, and an interval determining unit 1003.
A first set determining unit 1001, configured to determine, according to the previous M review time intervals of each diagnosis item of the user, an nth review time interval set of each diagnosis item according to normal distribution through data fitting; the set of nth review time intervals for each treatment item includes an nth review time interval for the historical user for each treatment item.
The first estimation determining unit 1002 is configured to determine an nth review time interval mean and an nth review time interval variance of an nth review time interval set of each diagnosis and treatment item by using the maximum likelihood estimation.
The interval determining unit 1003 is configured to determine an nth review time interval of each diagnosis item of the user according to the nth review time interval mean and the nth review time interval variance.
In the embodiment of the present invention, the first set determining unit 1001 determines, through data fitting, an nth review time interval set of each diagnosis item according to normal distribution, and further the first estimation determining unit 1002 determines, by using maximum likelihood estimation, an nth review time interval mean value and an nth review time interval variance of the nth review time interval set of each diagnosis item, and the final interval determining unit 1003 determines, based on the nth review time interval mean value and the nth review time interval variance, an nth review time interval of each diagnosis item of the user, so that accuracy of review interval prediction can be further improved.
Fig. 11 shows a schematic structure of an interval determining unit 1003 in a diagnosis and treatment prediction apparatus according to an embodiment of the present invention, and for convenience of explanation, only a portion related to the embodiment of the present invention is shown, and the details are as follows:
in an embodiment of the present invention, in order to improve the flexibility of determining the nth review time interval, referring to fig. 11, each unit included in the interval determining unit 1003 is configured to execute each step in the corresponding embodiment of fig. 4, and specifically refer to fig. 4 and the related description in the corresponding embodiment of fig. 4, which are not repeated herein. In the embodiment of the present invention, the interval determining unit 1003 includes a target interval determining subunit 1101 and an interval determining subunit 1102.
The target interval determination subunit 1101 is configured to determine a target review time interval according to the nth review time interval mean and the nth review time interval variance.
The interval determination subunit 1102 is configured to take the target review time interval as the nth review time interval of each diagnosis item of the user when the probability that the nth review time interval of each diagnosis item of the user falls into the target review time interval is not less than the preset probability value.
In the embodiment of the present invention, the target interval determining subunit 1101 determines the target review time interval according to the average value of the nth review time interval and the variance of the nth review time interval, and when the probability that the nth review time interval of each diagnosis item of the user falls into the target review time interval is not less than the preset probability value, the target review time interval is used as the nth review time interval of each diagnosis item of the user, and the flexibility of determining the nth review time interval can be improved by flexibly setting the preset probability value.
Fig. 12 shows another functional module of the diagnosis and treatment prediction apparatus according to the embodiment of the present invention, and for convenience of explanation, only the portion relevant to the embodiment of the present invention is shown in detail as follows:
In one embodiment of the invention, the user's medical cycle representation further includes the first M review costs for each medical item of the user. In order to improve accuracy of the review cost prediction, referring to fig. 12, each module included in the diagnosis and treatment prediction apparatus is configured to execute each step in the corresponding embodiment of fig. 5, and detailed descriptions of the steps in fig. 5 and the corresponding embodiment of fig. 5 will be omitted herein. In this embodiment of the present invention, based on the above-mentioned module structure, the diagnosis and treatment prediction apparatus further includes a cost determination module 1201.
The cost determining module 1201 is configured to determine, according to the first M review costs of each diagnosis and treat item of the user and the historical review costs of the historical user of each diagnosis and treat item, the nth review cost of each diagnosis and treat item of the user by using data fitting and maximum likelihood estimation.
In the embodiment of the invention, the cost determination module 1201 determines the nth review cost of each diagnosis item of the user by using data fitting and maximum likelihood estimation according to the previous M review costs of each diagnosis item of the user and the historical review costs of the historical user of each diagnosis item, so that accuracy of review cost prediction can be improved.
Fig. 13 shows a schematic structure of a cost determining module 1201 in the diagnosis and treatment prediction apparatus according to the embodiment of the present invention, and for convenience of explanation, only the portions related to the embodiment of the present invention are shown, which is described in detail below:
In an embodiment of the present invention, in order to improve accuracy of review cost prediction, referring to fig. 13, each unit included in the cost determining module 1201 is configured to perform each step in the corresponding embodiment of fig. 6, and detailed descriptions of the corresponding embodiments of fig. 6 are omitted herein. In the embodiment of the present invention, the cost determining module 1201 includes a second set determining unit 1301, a second estimation determining unit 1302, and a cost determining unit 1303.
A second set determining unit 1301, configured to determine, according to the first M review costs of each diagnosis item of the user, an nth review cost set of each diagnosis item according to normal distribution through data fitting; the nth review cost set for each treatment item includes an nth review cost for the historical user for each treatment item.
A second estimation determination unit 1302, configured to determine an nth review cost mean and an nth review cost variance of an nth review cost set of each diagnosis and treatment item using the maximum likelihood estimation;
the cost determining unit 1303 is configured to determine an nth review cost of each diagnosis and treatment item of the user according to the nth review cost mean and the nth review cost variance.
In the embodiment of the present invention, the second set determining unit 1301 determines, according to the previous M times of review cost of each diagnosis and treatment item of the user, the nth time of review cost set of each diagnosis and treatment item according to the normal distribution by data fitting, and further the second estimation determining unit 1302 determines the nth time of review cost mean and the nth time of review cost variance of the nth time of review cost set of each diagnosis and treatment item by using the maximum likelihood estimation, and the final cost determining unit 1303 determines the nth time of review cost of each diagnosis and treatment item of the user according to the nth time of review cost mean and the nth time of review cost variance, so that accuracy of review cost prediction can be improved.
Fig. 14 shows a schematic structure of the fee determining unit 1303 in the diagnosis and treatment predicting apparatus according to the embodiment of the present invention, and for convenience of explanation, only the portions related to the embodiment of the present invention are shown, and the details are as follows:
in an embodiment of the present invention, in order to improve the flexibility of review cost prediction, referring to fig. 14, each sub-unit included in the cost determining unit 1303 is configured to perform each step in the corresponding embodiment of fig. 7, and detailed descriptions in fig. 7 and the corresponding embodiment of fig. 7 are omitted herein. In the embodiment of the present invention, the fee determining unit 1303 includes a target fee determining subunit 1401 and a fee determining subunit 1402.
The target cost determination subunit 1401 is configured to determine a target review cost interval according to the nth review cost average value and the nth review cost variance.
The cost determination subunit 1402 is configured to take the target review cost interval as the nth review cost of each diagnosis item of the user when the probability that the nth review cost of each diagnosis item of the user falls into the target review cost interval is not less than the preset probability value.
In the embodiment of the present invention, the target cost determination subunit 1401 determines a target review cost interval according to the nth review cost average value and the nth review cost variance, and when the probability that the nth review cost of each diagnosis item of the user falls into the target review cost interval is not less than the preset probability value, the target review cost interval is used as the nth review cost of each diagnosis item of the user, so that the flexibility of review cost prediction can be improved.
In an embodiment of the present invention, in order to implement prediction of total diagnosis and treatment cost of each diagnosis and treatment item of the user, the diagnosis and treatment period portrait of the user further includes a history diagnosis and treatment cost of each diagnosis and treatment item of the user. On the basis of the above-mentioned modular structure, diagnose prediction device still includes:
The total cost prediction module is used for determining the total cost of diagnosis and treatment of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation according to the calendar diagnosis and treatment cost of each diagnosis and treatment item of the user and the historical total cost of diagnosis and treatment of the historical user of each diagnosis and treatment item.
In one embodiment of the present invention, to improve accuracy of total cost prediction for diagnosis and treatment, the total cost prediction module includes:
the third set determining unit is used for determining a diagnosis and treatment total cost set of each diagnosis and treatment item according to the history diagnosis and treatment cost of each diagnosis and treatment item of the user through data fitting; the aggregate total cost per treatment item includes the total cost per treatment item for the historic user's treatment.
And the third estimation determining unit is used for determining the diagnosis and treatment total cost mean value and the diagnosis and treatment total cost variance of the diagnosis and treatment total cost set of each diagnosis and treatment item by using the maximum likelihood estimation.
The total cost determining unit is used for determining the diagnosis and treatment total cost of each diagnosis and treatment item of the user according to the diagnosis and treatment total cost mean value and the diagnosis and treatment total cost variance.
In one embodiment of the present invention, in order to improve flexibility of diagnosis and treatment total cost prediction, the total cost determination unit includes:
the target total cost determining subunit is used for determining a target total cost interval according to the total cost mean value and the total cost variance.
The total cost determination subunit is configured to use the target total cost diagnosis interval as the total cost of diagnosis and treatment of each diagnosis and treatment item of the user when the probability that the total cost of diagnosis and treatment of each diagnosis and treatment item of the user falls into the target total cost diagnosis and treatment interval is not less than the preset probability value.
The following specifically describes the principles and processes of diagnosis and treatment prediction provided by the embodiment of the present invention, taking dental diagnosis and treatment as an example:
(1) Acquiring consumption medical data of each tooth of a user, namely a diagnosis and treatment record of each tooth of the user;
(2) Preprocessing the diagnosis and treatment record of each tooth of a user, such as desensitization treatment of customer sensitive information, null value or missing data treatment, special character conversion, repeated data duplication removal, inconsistent data merging, data digitizing and discretizing, medical diagnosis text analysis, classification and structuring treatment and the like;
(3) Extracting information in the diagnosis and treatment record of each tooth of the user after treatment, such as information of keywords, initial diagnosis time points, re-diagnosis time points, diagnosis and treatment cost and the like;
(4) Determining at least one diagnosis and treatment item of each tooth through the extracted keywords, and determining the time interval, the cost and the like of the last re-diagnosis and treatment item;
the following provides a partial keyword diagnosis and treatment item data table for an exemplary dental diagnosis and treatment for reference:
Partial keyword diagnosis and treatment item data table for surface two-tooth diagnosis and treatment
(5) And generating a diagnosis and treatment period image of the teeth of the user by using the at least one diagnosis and treatment item, and the information such as the time interval of the last re-diagnosis and the cost of the last re-diagnosis of each diagnosis and treatment item.
An exemplary diagnostic periodic representation of a user's teeth is provided below for reference:
medical period portrait of three-purpose user teeth
From the above table, it can be seen that, for the diagnosis and treatment item of "resin dental filling", the user makes a first diagnosis at time point T20, and the cost of diagnosis and treatment is a20; further, the first review is performed at time T21, with the first review being at cost a21 and the first review (with the initial visit) being at time interval B21.
Aiming at the diagnosis and treatment item of 'dental implant', the user makes a first diagnosis at a time point T40, and the cost of diagnosis and treatment is A40; the first review is performed at time T41, the first review costs a fee a41, the first review (with the first review) takes a time interval B41, the second review is performed at time T42, the second review costs a42, the second review (with the first review or with the first review) takes a time interval B42, the third review is performed at time T43, the third review costs a43, and the third review (with the first review or with the second review) takes a time interval B43; at time T4 (N-2), the (N-2) th review was performed at a cost of A4 (N-2) and at a time interval of B4 (N-2) with the (N-2) th review (with the initial or with the (N-3) th review).
The diagnosis and treatment items of whitening and repairing, painless tooth extraction, tooth orthodontic and the like are similar to the diagnosis and treatment items of resin tooth filling and tooth planting, and are not illustrated one by one.
(6) And respectively aiming at each diagnosis and treatment item, determining the Nth review time interval of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis and treatment item and the historical review time interval of the historical user of each diagnosis and treatment item.
For example, taking the diagnosis and treatment item "dental implant" as an example, according to the previous M times of review time intervals of the diagnosis and treatment item "dental implant" of the user, determining the nth time of review time interval set of the diagnosis and treatment item "dental implant" conforming to normal distribution through data fitting; the nth review time interval set of the dental implant diagnosis and treatment item comprises the nth review time interval of the historical user of the dental implant diagnosis and treatment item; and determining an nth re-diagnosis time interval mean value and an nth re-diagnosis time interval variance of an nth re-diagnosis time interval set of the 'dental implant' diagnosis and treatment item by using the maximum likelihood estimation, and taking the target re-diagnosis time interval as the nth re-diagnosis time interval of the 'dental implant' diagnosis and treatment item of the user when the probability that the nth re-diagnosis time interval of the 'dental implant' diagnosis and treatment item of the user falls into the target re-diagnosis time interval is not smaller than a preset probability value.
If the user set is denoted as { C k K=1, 2..a.the user's diagnosis and treatment item set is noted as { Treat m M=1, 2,..m. For a certain user C k Treat in its diagnosis and treatment item m In one diagnosis and treatment period of (1), the date of the first diagnosis is recorded as Day 0 (Treat m |C k ) The date of the first review is recorded as Day 1 (Treat m |C k ) The date of the last review is recorded as Day N (Treat m |C k ). Treat in this diagnosis and treatment item m In the diagnosis and treatment period of (1), all diagnosis and treatment dates are marked as { Day } n (Treat m |C k ) N=1,..n. Thus, user C k Treat of diagnosis and treatment item m Future diagnosis and treatment date Day of (2) n (Treat m |C k ) Can be expressed as a functional relationship ofWhere t represents some point in time in the future.
When the time interval of the re-diagnosis is the interval between the next re-diagnosis and the previous re-diagnosis, the date interval of the first re-diagnosis is recorded as Period 1 (Treat m |C k )＝Day 1 (Treat m |C k )-Day 0 (Treat m |C k ). The date interval between the nth and the 1 st re-diagnosis is recorded as Period N (Treat m |C k )＝Day N (Treat m |C k )-Day N-1 (Treat m |C k ). The date interval between the date of the Nth review and the date of the first review is marked as Period N (Treat m |C k )＝Day N (Treat m |C k )-Day 0 (Treat m |C k )。
For all historical users of the diagnosis and treatment item, the treatment item Treat m In the diagnosis and treatment period of the history user, the nth re-diagnosis time interval of the history user is determined by utilizing the nth re-diagnosis time and the (N-1) th re-diagnosis time of the history user on the assumption that the diagnosis and treatment record of the history user contains the nth re-diagnosis record of the history user. And if the diagnosis and treatment record of the historical user does not contain the Nth review record of the historical user, obtaining the Nth review time of the historical user through data fitting, and determining the Nth review time interval of the historical user by utilizing the Nth review time and the (N-1) th review time of the historical user obtained through fitting.
After determining the nth review time interval of all historical users, the nth review date interval Period of all previous users n (Treat m ) Is approximately subjected to normal distribution N (mu) m,n ,σ m,n 2 ). The average value of the nth re-diagnosis time interval is recorded asThe mean and variance of the normal distribution are estimated by a maximum likelihood estimation method to obtain a normal distribution N (mu) m,n ,σ m,n 2 ) Is +.>Variance is
Because of the fact that,therefore, for a certain user C k In other words, it diagnoses item Treat m The time interval of the future nth review of (2) falls approximately within the interval +.>The probability of (2) was 95%. Thus, user C k Treat of diagnosis and treatment item m The future diagnosis and treatment time interval of (2) can be expressed as +.>
(7) And respectively aiming at each diagnosis and treatment item, determining the Nth review cost of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review costs of each diagnosis and treatment item of the user and the historical review cost of the historical user of each diagnosis and treatment item.
For example, taking the diagnosis item "dental implant" as an example, according to the previous M times of review cost of the diagnosis item "dental implant" of the user, determining an nth time of review cost set of the diagnosis item "dental implant" according to normal distribution through data fitting, wherein the nth time of review cost set of the diagnosis item "dental implant" comprises the nth time of review cost of the historical user of the diagnosis item "dental implant". And determining an nth review cost mean value and an nth review cost variance of an nth review cost set of the dental implant diagnosis and treatment item by using the maximum likelihood estimation, and taking the target review cost interval as the nth review cost of the dental implant diagnosis and treatment item when the probability that the nth review cost of the dental implant diagnosis and treatment item falls into the target review cost interval is not smaller than a preset probability value.
Similarly, for a certain user C k Treat in its diagnosis and treatment item m In a diagnosis and treatment period of (1), the initial diagnosis Cost is recorded as Cost 1 (Treat m |C k ) The first review Cost is recorded as Cost 2 (Treat m |C k ) ,. the last review Cost is recorded as Cost N (Treat m |C k ). For all historical users of the diagnosis and treatment item, the treatment item Treat m In the diagnosis and treatment period of the history user, assuming that the diagnosis and treatment record of the history user contains the nth review cost of the history user, taking the nth review cost of the history user as the nth review cost of the history user; suppose a history of diagnosis and treatment of a userAnd if the historical user's nth review cost is not included, obtaining the historical user's nth review cost through data fitting, and using the historical user's nth review cost obtained through fitting as the historical user's nth review cost.
After determining the nth review costs for all the historical users, the nth review costs for all the historical users are approximately subject to a normal distribution. And estimating the mean value and the variance of the normal distribution by a maximum likelihood estimation method. Obtaining the average valueVariance->After that, for user C k Its diagnosis and treatment item Treat m The nth review cost of (2) falls approximately in the interval +.>The probability of (2) was 95%.
Similarly, for a certain user C k Treat in its diagnosis and treatment item m In a diagnosis and treatment period of (1), the initial diagnosis Cost is recorded as Cost 1 (Treat m |C k ) The first review Cost is recorded as Cost 2 (Treat m |C k ) ,. the last review Cost is recorded as Cost N (Treat m |C k ). In the diagnosis and treatment period of the diagnosis and treatment item, the total diagnosis and treatment cost is recorded asFor all historical users, the treatment item Treat is used for m In a diagnosis and treatment period of (2), the average diagnosis and treatment total cost is recorded as
For all historical users of the diagnosis and treatment item, the treatment item Treat m In the diagnosis and treatment period of the history user, the diagnosis and treatment record of the history user comprises the diagnosis and treatment total Cost (Treat) of the history user m ) Diagnosis and treatment item Treat of all historical users m Cost (Treat) of total diagnosis and treatment m ) Is approximately subjected to normal distribution N (mu) n ,σ n 2 ). For this normal distribution N (mu) by maximum likelihood estimation n ,σ n 2 ) The mean and variance of (c) are estimated. The average value is obtained asVariance isBecause of (I)>Therefore, for user C k Its diagnosis and treatment item Treat m Is the estimated total Cost of diagnosis and treatment pred (Treat m |C k ) Falls approximately in the interval->The probability of (2) was 95%.
The embodiment of the invention also provides computer equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the diagnosis and treatment prediction method is realized when the processor executes the computer program.
The embodiment of the invention also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program for executing the diagnosis and treatment prediction method.
In summary, in the embodiment of the present invention, first, the consumption medical data of the user is obtained, and then the diagnosis and treatment period portrait of the user is determined according to the consumption medical data of the user; and finally, determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item. According to the embodiment of the invention, the diagnosis and treatment period portrait of the user is determined through consuming the medical data, and then the Nth review time interval of each diagnosis and treatment item of the user is determined through data fitting and maximum likelihood estimation, so that the accuracy of review interval prediction can be improved.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.

Claims (7)

1. A diagnosis and treatment prediction method, comprising:
acquiring consumption medical data of a user;
determining a diagnosis and treatment period portrait of the user according to the consumption medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point;
Determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item;
wherein M is a natural number, N is a positive integer, and M is smaller than N;
according to the previous M times of review time intervals of each diagnosis and treatment item of the user and the historical review time intervals of the historical user of each diagnosis and treatment item, determining the nth review time interval of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation, wherein the method comprises the following steps:
according to the previous M times of re-diagnosis time intervals of each diagnosis and treatment item of the user, determining an nth time of re-diagnosis time interval set of each diagnosis and treatment item conforming to normal distribution through data fitting; the nth review time interval set of each diagnosis item comprises an nth review time interval of a historical user of each diagnosis item;
determining an nth re-diagnosis time interval mean value and an nth re-diagnosis time interval variance of an nth re-diagnosis time interval set of each diagnosis and treatment item by using maximum likelihood estimation;
determining a target re-diagnosis time interval according to the average value of the nth re-diagnosis time interval and the variance of the nth re-diagnosis time interval;
When the probability that the nth re-diagnosis time interval of each diagnosis and treatment item of the user falls into the target re-diagnosis time interval is not smaller than a preset probability value, taking the target re-diagnosis time interval as the nth re-diagnosis time interval of each diagnosis and treatment item of the user;
determining a medical cycle representation of the user based on the consumer medical data of the user, comprising:
extracting keywords in the user consumption medical data, and determining each diagnosis and treatment item of the user according to the extracted keywords in the user consumption medical data: after extracting keywords in the medical data consumed by the user, determining diagnosis and treatment items corresponding to the keywords from a keyword diagnosis and treatment item data table, wherein the keyword diagnosis and treatment item data table comprises the keywords, the diagnosis and treatment items and the corresponding relation between the keywords and the diagnosis and treatment items, and the keywords corresponding to the diagnosis and treatment items in the keyword diagnosis and treatment item data table comprise one or more keywords;
extracting the diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data, and determining the previous M times of re-diagnosis time intervals of each diagnosis and treatment item according to the diagnosis and treatment time of each diagnosis and treatment item in the extracted user consumption medical data;
and determining the diagnosis and treatment period portrait of the user according to the previous M times of the review time interval of each diagnosis and treatment item of the user.
2. The method of claim 1, wherein the representation of the user's treatment cycle further comprises a first M review costs for each treatment item of the user, the method of predicting the treatment further comprising:
and determining the Nth review cost of each diagnosis and treatment item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review costs of each diagnosis and treatment item of the user and the historical review costs of the historical user of each diagnosis and treatment item.
3. The method of claim 2, wherein determining the nth review cost for each of the user's diagnosis items using data fitting and maximum likelihood estimation based on the first M review costs for each of the user's diagnosis items and the historical review costs for each of the diagnosis items, comprising:
according to the previous M times of review cost of each diagnosis and treatment item of the user, determining an nth review cost set of each diagnosis and treatment item conforming to normal distribution through data fitting; the nth review cost set of each treatment item includes an nth review cost of the historical user of each treatment item;
determining an nth review cost mean value and an nth review cost variance of an nth review cost set of each diagnosis and treatment item by using the maximum likelihood estimation;
And determining the nth re-diagnosis cost of each diagnosis and treatment item of the user according to the nth re-diagnosis cost mean value and the nth re-diagnosis cost variance.
4. The method of claim 3, wherein determining the nth review cost for each of the user's categories of treatment based on the nth review cost mean and the nth review cost variance comprises:
determining a target review cost interval according to the Nth review cost mean value and the Nth review cost variance;
when the probability that the nth review cost of each diagnosis and treatment item of the user falls into the target review cost interval is not smaller than a preset probability value, the target review cost interval is used as the nth review cost of each diagnosis and treatment item of the user.
5. A diagnosis and treatment prediction apparatus, comprising:
the data acquisition module is used for acquiring the consumption medical data of the user;
the portrait determining module is used for determining the portrait of the diagnosis and treatment period of the user according to the consumption medical data of the user; the diagnosis and treatment period portrait of the user comprises the first M times of re-diagnosis time intervals of each diagnosis and treatment item of the user; the re-diagnosis time interval is the time interval between the last diagnosis time point and the previous diagnosis time point, or the time interval between the current diagnosis time point and the initial diagnosis time point;
The interval determining module is used for determining the Nth review time interval of each diagnosis item of the user by utilizing data fitting and maximum likelihood estimation according to the previous M review time intervals of each diagnosis item of the user and the historical review time intervals of the historical user of each diagnosis item;
wherein M is a natural number, N is a positive integer, and M is smaller than N;
an interval determination module comprising:
the first set determining unit is used for determining an nth review time interval set of each diagnosis and treatment item according to the previous M review time intervals of each diagnosis and treatment item of the user through data fitting; the nth review time interval set of each diagnosis item comprises an nth review time interval of a historical user of each diagnosis item;
the first estimation determining unit is used for determining an nth re-diagnosis time interval mean value and an nth re-diagnosis time interval variance of an nth re-diagnosis time interval set of each diagnosis and treatment item by using the maximum likelihood estimation;
the interval determining unit is used for determining a target review time interval according to the average value of the N-th review time interval and the variance of the N-th review time interval; when the probability that the nth re-diagnosis time interval of each diagnosis and treatment item of the user falls into the target re-diagnosis time interval is not smaller than a preset probability value, taking the target re-diagnosis time interval as the nth re-diagnosis time interval of each diagnosis and treatment item of the user;
The portrait determination module includes:
the keyword extraction unit is used for extracting keywords in the user consumption medical data and determining each diagnosis and treatment item of the user according to the extracted keywords in the user consumption medical data: after extracting keywords in the medical data consumed by the user, determining diagnosis and treatment items corresponding to the keywords from a keyword diagnosis and treatment item data table, wherein the keyword diagnosis and treatment item data table comprises the keywords, the diagnosis and treatment items and the corresponding relation between the keywords and the diagnosis and treatment items, and the keywords corresponding to the diagnosis and treatment items in the keyword diagnosis and treatment item data table comprise one or more keywords;
the diagnosis and treatment time extraction unit is used for extracting the diagnosis and treatment time of each diagnosis and treatment item in the user consumption medical data and determining the previous M times of re-diagnosis time intervals of each diagnosis and treatment item according to the diagnosis and treatment time of each diagnosis and treatment item in the extracted user consumption medical data;
and the portrait determining unit is used for determining the portrait of the diagnosis and treatment period of the user according to each diagnosis and treatment item of the user and the previous M times of review time intervals of each diagnosis and treatment item.
6. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the diagnosis and treatment prediction method of any one of claims 1 to 4 when the computer program is executed by the processor.
7. A computer-readable storage medium storing a computer program for executing the diagnosis and treatment prediction method according to any one of claims 1 to 4.
CN202011030257.3A 2020-09-27 2020-09-27 Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium Active CN112133429B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011030257.3A CN112133429B (en) 2020-09-27 2020-09-27 Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011030257.3A CN112133429B (en) 2020-09-27 2020-09-27 Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium

Publications (2)

Publication Number Publication Date
CN112133429A CN112133429A (en) 2020-12-25
CN112133429B true CN112133429B (en) 2023-12-22

Family

ID=73840519

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011030257.3A Active CN112133429B (en) 2020-09-27 2020-09-27 Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium

Country Status (1)

Country Link
CN (1) CN112133429B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09259177A (en) * 1996-03-19 1997-10-03 Fujitsu General Ltd Medical re-examination reservation system
JP2005202686A (en) * 2004-01-15 2005-07-28 Nooza:Kk Medical expense estimation system, medical expense estimation method and medical expense estimation program
JP2013045419A (en) * 2011-08-26 2013-03-04 Chugoku Electric Power Co Inc:The Service time prediction device and service time prediction method
KR20140066499A (en) * 2012-11-23 2014-06-02 (주)솔트웍스 Clinical decision support system devices and method thereof
CN106874653A (en) * 2017-01-13 2017-06-20 深圳市前海安测信息技术有限公司 Assistant hospital decision system and method
EP3573068A1 (en) * 2018-05-24 2019-11-27 Siemens Healthcare GmbH System and method for an automated clinical decision support system

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050119534A1 (en) * 2003-10-23 2005-06-02 Pfizer, Inc. Method for predicting the onset or change of a medical condition

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09259177A (en) * 1996-03-19 1997-10-03 Fujitsu General Ltd Medical re-examination reservation system
JP2005202686A (en) * 2004-01-15 2005-07-28 Nooza:Kk Medical expense estimation system, medical expense estimation method and medical expense estimation program
JP2013045419A (en) * 2011-08-26 2013-03-04 Chugoku Electric Power Co Inc:The Service time prediction device and service time prediction method
KR20140066499A (en) * 2012-11-23 2014-06-02 (주)솔트웍스 Clinical decision support system devices and method thereof
CN106874653A (en) * 2017-01-13 2017-06-20 深圳市前海安测信息技术有限公司 Assistant hospital decision system and method
EP3573068A1 (en) * 2018-05-24 2019-11-27 Siemens Healthcare GmbH System and method for an automated clinical decision support system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
口腔医院信息系统的设计与实现;蒋冰 等;《中国优秀硕士学位论文全文数据库 信息科技辑》(第04期);第1-73页 *
门诊护士站关于优化口腔内科患者就诊流程的研究;张玉英 等;《中国卫生标准管理》(第21期);第168-170页 *

Also Published As

Publication number Publication date
CN112133429A (en) 2020-12-25

Similar Documents

Publication Publication Date Title
US11922348B2 (en) Generating final abnormality data for medical scans based on utilizing a set of sub-models
US5893069A (en) System and method for testing prediction model
US6061657A (en) Techniques for estimating charges of delivering healthcare services that take complicating factors into account
US9443002B1 (en) Dynamic data analysis and selection for determining outcomes associated with domain specific probabilistic data sets
US20120271612A1 (en) Predictive modeling
US20140006044A1 (en) System and method for preparing healthcare service bundles
CN112562836A (en) Doctor recommendation method and device, electronic equipment and storage medium
CN114496264B (en) Health index analysis method, device, equipment and medium based on multidimensional data
CN112016905B (en) Information display method and device based on approval process, electronic equipment and medium
CN113657548A (en) Medical insurance abnormity detection method and device, computer equipment and storage medium
CN113164163A (en) Device, method and program for inferring depression state
CN112133429B (en) Diagnosis and treatment prediction method and device, computer equipment and computer readable storage medium
Gustafson et al. Subjective probabilities in medical diagnosis
JP6481794B1 (en) Learning data generation method, learning data generation program
CN115910387A (en) Data processing method, device and equipment based on time sequence and storage medium
CN116312926A (en) Health path recommending method and related device, electronic equipment and storage medium
CN111460293B (en) Information pushing method and device and computer readable storage medium
CN110517779B (en) Data processing method and device, electronic equipment and readable storage medium
Rodriguez User interface to automate the collection and processing of data for discrete event simulation projects
US11631494B2 (en) System and method for monitoring installed healthcare assets
CN116957138A (en) Diagnosis and treatment related cost prediction method, system, medium and terminal
Padmasawitri et al. The Influence of Modelling Assumptions on the Dynamic Transmission Model’s Cost-Effectiveness Projection of Tuberculosis Intervention
CN116884587A (en) Method and device for analyzing high-value consumable material cost of operation
Mesaros et al. HTA Analysis of Cochlear Implant-Health Technology Assessment in Slovak Healthcare Environment
CN117292841A (en) Disease risk factor selection method and system based on evidence-based medicine

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant