CN111477321A - Treatment effect prediction system with self-learning capability and treatment effect prediction terminal - Google Patents
Treatment effect prediction system with self-learning capability and treatment effect prediction terminal Download PDFInfo
- Publication number
- CN111477321A CN111477321A CN202010167791.2A CN202010167791A CN111477321A CN 111477321 A CN111477321 A CN 111477321A CN 202010167791 A CN202010167791 A CN 202010167791A CN 111477321 A CN111477321 A CN 111477321A
- Authority
- CN
- China
- Prior art keywords
- treatment effect
- case data
- sample
- module
- prediction
- 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.)
- Granted
Links
Images
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Epidemiology (AREA)
- Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application provides a treatment effect prediction system with self-learning capability and a treatment effect prediction terminal, and relates to the technical field of diagnosis and treatment instruments. The method aims to solve the problem that a treatment effect prediction instrument can only predict the treatment effect by using an original prediction model. The case data acquisition module is used for receiving and inputting new case data; the treatment effect prediction module is used for obtaining a predicted treatment effect by using a prediction model according to the new case data and sending the predicted treatment effect to the treatment effect display module; the case data classification module is used for obtaining a target treatment effect according to the target classification of the new case data and taking the target treatment effect as a label of the new case data; the database is used for adding new case data with labels to obtain a sample data set; the treatment effect prediction module is also used for training and updating the prediction model by the sample data set and predicting the next treatment effect by using the updated prediction model.
Description
Technical Field
The application relates to the technical field of diagnosis and treatment instruments, in particular to a treatment effect prediction system with self-learning capability and a treatment effect prediction terminal.
Background
Recently, machine learning algorithms are increasingly used in the fields of education, traffic, finance, and the like. Medical fields are also actively being explored for assisting diagnosis, examination, and the like, using machine learning techniques. Taking evaluation of the postoperative curative effect of nasosinusitis as an example, most researches adopt a logistic regression algorithm in machine learning to analyze risk factors influencing the postoperative curative effect of nasosinusitis, so that clinicians can be helped to comprehensively consider and make more effective prevention and control measures, but the algorithm models have the following disadvantages:
1. once the algorithm model is trained, the model does not have the capability of automatic upgrading in the future use process.
2. Medical workers can only obtain the results of the algorithms in the process of using the algorithms, and cannot interact with the algorithms.
Disclosure of Invention
In view of the above problems, embodiments of the present application provide a treatment effect prediction system and a treatment effect prediction terminal with self-learning capability, which aim to solve the problem that a treatment effect prediction apparatus can only predict a treatment effect with an original prediction model.
A first aspect of an embodiment of the present application provides a therapeutic effect prediction system, including: the system comprises a case data acquisition module, a treatment effect prediction module, a case data classification module, a treatment effect display module and a database; the case data acquisition module is connected with the treatment effect prediction module and the case data classification module;
the case data acquisition module is used for receiving and inputting new case data;
the treatment effect prediction module is used for obtaining a predicted treatment effect by using a prediction model according to the new case data and sending the predicted treatment effect to the treatment effect display module;
the case data classification module is used for obtaining a target treatment effect according to the target classification of the new case data and taking the target treatment effect as a label of the new case data;
the database is used for adding new case data with labels to obtain a sample data set;
the treatment effect prediction module is also used for training and updating the prediction model by the sample data set and predicting the next treatment effect by using the updated prediction model.
Optionally, the system further comprises:
a labeling module for labeling an initial label for the new case data, the initial label being the predicted treatment effect;
and the label updating module is used for replacing the predicted treatment effect with the target treatment effect and updating the label of the new case data when the predicted treatment effect is different from the target treatment effect.
Optionally, the system further comprises:
and the prompting module is used for outputting a prompting message to prompt a user to verify the label of the new case data when the predicted treatment effect is different from the target treatment effect.
Optionally, the system further comprises a verification module;
the checking module is used for determining the label of the new case data according to the recorded actual treatment effect;
wherein the recorded therapeutic effect is the target therapeutic effect or the predicted therapeutic effect or a therapeutic effect other than the target therapeutic effect and the predicted therapeutic effect;
when the recorded treatment effect is other treatment effect, the database takes the recorded treatment effect as the label of the new case data and updates the sample data set;
the treatment effect prediction module is also used for training and updating the prediction model by using the updated sample data set, and predicting the treatment effect of the next time by using the updated prediction model.
Optionally, the database stores an original sample data set;
the case data classification module comprises a merging submodule, a clustering submodule and a classification submodule;
the merging submodule is used for acquiring the original sample data set from the database and merging the new case data and the original sample data set to obtain a sample set to be classified;
the clustering submodule is used for clustering the sample set to be classified to obtain a plurality of sample classifications;
the classification submodule is used for determining the sample classification where the new case data is located as the target classification.
Optionally, a single sample in the original sample data set is original case data carrying a treatment effect label;
the merging submodule comprises a merging subunit;
the merging subunit is configured to merge the new case data and the samples of the original sample data set according to the original case data and the treatment effect label of each sample in the original sample data set, and the new case data and the predicted treatment effect.
Optionally, the case data classification module further comprises a statistics submodule and a labeling submodule;
the statistic submodule is used for counting the number of case data with the same treatment effect in the target classification;
the labeling submodule is configured to label the treatment effect with the largest number of case data as the target treatment effect.
Optionally, the clustering sub-module further comprises a labeling sub-unit for labeling all case data in the sample classification with the sample classification;
the clustering submodule comprises a clustering subunit, and the clustering subunit is used for taking the sample classification marked by the new case data as a first sample classification and determining all case data included in the first sample classification as a target classification; wherein all case data comprised by the first sample classification belong to the set of samples to be classified.
Optionally, the system further comprises:
the communication module is used for receiving a modification instruction for modifying the samples in the original sample data set;
the database is also used for updating the original sample data set according to the modification instruction;
the treatment effect prediction module is also used for training and updating the prediction model by the updated original sample data set when the original sample data set is updated, and predicting the treatment effect of the next time by using the updated prediction model.
A second aspect of the embodiments of the present application provides a therapeutic effect prediction terminal, where the therapeutic effect prediction terminal is an integrated terminal of part or all of modules in the therapeutic effect prediction system described in the first aspect;
the treatment effect prediction terminal also comprises a display screen and an information input component, wherein the display screen is used for displaying the data output by the treatment effect prediction system; the information input component is used for a user to input data to the treatment effect prediction system.
In summary, the treatment effect prediction system provided in the embodiment of the present application is provided with a case data classification module, and clusters input new case data with a large amount of original case data locally stored in the treatment effect prediction system, so that the treatment effect prediction system can predict the treatment effect of the new case data, and can obtain a target treatment effect that the new case data should have compared with other case data based on a clustering algorithm, thereby obtaining an empirical treatment effect conforming to other case data; then, the obtained predicted treatment effect is checked, and the prediction model is updated by the new case data with the target treatment effect label, so that the treatment effect prediction system has the error correction capability based on the new case data sample and the self-learning capability of continuously optimizing the original prediction model; meanwhile, the treatment effect prediction system provided by the embodiment of the application is provided with the verification module, when the prediction treatment effect is inaccurate, the medical staff is reminded to verify the prediction treatment effect, and the prediction model is updated according to the verification result of the medical staff (namely the actual treatment effect of the patient input by the medical staff), so that the treatment effect prediction system provided by the embodiment of the application has an interactive function of receiving feedback of the medical staff.
In addition, based on the treatment effect prediction system of the embodiment of the application, medical staff can actively modify case data in the treatment effect prediction system, update the prediction model with more accurate case data after modification, further enhance the model prediction accuracy of the treatment effect prediction system, and meanwhile, the treatment effect prediction system can automatically modify the label of new case data when receiving the new case data, also provides a function of actively modifying the total case data of the system, and improves the sample error correction capability and the interaction performance of the treatment effect prediction system.
In addition, the invention provides a nasosinusitis treatment effect prediction terminal with self-learning capability. According to the invention, by means of the information acquisition terminal, the relevant data (including numerical values and text data) of the medical history data of the patient is fed back to the system in real time, the system automatically divides the samples through a clustering algorithm, and meanwhile, experienced medical workers can correct the wrongly divided samples in the system. When any new sample is collected, such as the system, the system can automatically learn the data again, and the algorithm is upgraded.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments of the present application will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive exercise.
FIG. 1 is a schematic structural diagram of a system for predicting therapeutic effect according to an embodiment of the present disclosure;
FIG. 2 is a schematic structural diagram of a case data classification module according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a system for predicting treatment effectiveness according to another embodiment of the present application;
fig. 4 is a schematic diagram of a result of the therapeutic effect prediction terminal according to the embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The treatment effect prediction instrument is provided with a prediction model, and the prediction model analyzes and calculates the case data based on a machine learning algorithm to obtain a predicted treatment effect, specifically, a neural network (such as L STM neural network, convolutional neural network and the like) can be used for learning the characteristics in the case data, machine learning methods such as supervised learning, unsupervised learning, reinforcement learning and the like can be used for training the model based on a large number of case data-treatment effect samples to obtain an initial prediction model capable of predicting the treatment effect of an illness based on the case data, and the method for obtaining the initial prediction model is not limited in the embodiment of the application.
The case data is all relevant information of the patient from the entrance of an outpatient clinic to the discharge of the patient with or without the patient. For example, age, height, white blood cell count in blood routine laboratory test reports, brain CT data, runny nose, cough, penicillin, acyclovir and the like can be used as case data.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a therapeutic effect prediction system according to an embodiment of the present application.
A first aspect of an embodiment of the present application provides a therapeutic effect prediction system, including: a case data acquisition module 1, a treatment effect prediction module 2, a case data classification module 3, a treatment effect display module 4 and a database 5; the case data acquisition module 1 is connected with the treatment effect prediction module 2 and the case data classification module 3;
the case data acquisition module 1 is used for receiving and inputting new case data;
the case data acquisition module 1 can receive case data entered by medical workers through an interactive page. For example, the treating physician enters the symptoms of nasal discharge, cough and the like and the treatment information of prescription penicillin and the like of the patient; the doctor in the laboratory enters information on the test order such as the blood routine.
The case data acquisition module 1 also converts new case data entered by medical workers into machine language.
The treatment effect prediction module 2 is used for obtaining a predicted treatment effect by using a prediction model according to the new case data and sending the predicted treatment effect to the treatment effect display module 4;
the treatment effect display module 4 can be connected with the treatment effect prediction module 2, or the treatment effect display module 4 is directly arranged in the treatment effect prediction module 2, so that the treatment effect display module 4 can be connected with the treatment effect prediction module 2 through the same interactive page, and after medical staff enters new case data, the predicted treatment effect of the new case data can be immediately obtained.
The therapeutic effect prediction module 2 is provided with a prediction model, which can be an initial prediction model or an updated prediction model obtained after a period of self-learning.
The prediction model can output the predicted treatment effect of the new case data corresponding to the disease according to the characteristics of the new case data received by the case data acquisition module 1. For example, the prediction model finds out that the current nasosinusitis case can be cured according to the chief complaint text information and the laboratory sheet information of a nasosinusitis case; the 'cure' is the treatment effect of the treatment effect prediction module 2 on the prediction of the nasosinusitis case by using the prediction model, the treatment effect prediction module 2 outputs the 'cure' prediction treatment effect information and sends the information to the treatment effect display module 4, and the treatment effect display module 4 displays the 'cure' prediction treatment effect on the interactive page of the treatment effect prediction system.
The treatment effect prediction system further comprises a labeling module 6 for labeling the new case data with an initial label, wherein the initial label is the predicted treatment effect;
for example, after the predicted therapeutic effect of the therapeutic effect prediction module 2 is "cured", the labeling module 6 stores "cured" as a label of a case of sinusitis in the therapeutic effect prediction system together with the case of sinusitis. Generally can be<sample,labelpred>In the form of<Data on cases of sinusitis, cure>Is stored. Sample is used to broadly refer to new case data.
The labeling module 6 can be a sub-module of the treatment effect prediction module 2, and after the treatment effect prediction module 2 outputs the predicted treatment effect, the predicted treatment effect is sent to the treatment effect display module 4 for informing medical workers of the predicted treatment effect, and the labeling module 6 labels the new case data, so that the storage form of the new case data is the same as that of the original sample data concentrated samples in the database 5, and the case data can be conveniently merged and classified subsequently.
The labeling module 6 may be a separate module connected to the therapeutic effect prediction module 2.
The case data classification module 3 is used for obtaining a target treatment effect according to the target classification of the new case data, and taking the target treatment effect as a label of the new case data;
the database 5 stores an original sample data set; a single sample in the original sample data set is original case data carrying a treatment effect label;
the original sample data set comprises a large number of samples, wherein the samples comprise samples for training an initial prediction model and samples which are historically received and stored by the treatment effect prediction system after the treatment effect prediction system starts working and the prediction model is updated for multiple times.
Samples in the original sample data set are all processed by<datai,labeli>Is stored. Where i refers to the value of the current sample in the original sample data set, for example when i is 5,<datai,labeli>refers to the centralized storage of original sample dataThe 5 th sample stored; when the total number of samples in the original sample data set is 100, the value of i is 1 to 100. The original sample Data set can also be regarded as the initial case database Datadb，<datai,labeli>Can be expressed as<Case dataiTherapeutic effectsi>In the form of (1).
Therapeutic effect of concentrated sample of original sample dataiIs case dataiActual treatment effect of corresponding disease condition or case data of medical worker to treatment effect prediction systemiAnd (4) verifying the corresponding disease condition predicted treatment effect to obtain the verified treatment effect.
Referring to fig. 2, fig. 2 is a schematic structural diagram of a case data classification module according to an embodiment of the present application.
The case data classification module 3 comprises a merging submodule 31, a clustering submodule 32 and a classification submodule 33;
the merging submodule 31 is configured to obtain the original sample data set from the database 5, and merge the new case data with the original sample data set to obtain a sample set to be classified;
the case data classification module 3 respectively obtains new case data with specific labels from the database 5<Data on cases of sinusitis, cure>And an original sample data set<Raw case dataiTherapeutic effectsi>. That is, the case data classification module 3 acquires from the database 5<sample,labelpred>And<datai,labeli>。
wherein new case data<sample,labelpred>I.e. by<Data on cases of sinusitis, cure>The database 5 may be directly sent to the case data classification module 3, or the merging submodule 31 may be obtaining the original sample data set<Raw case dataiTherapeutic effectsi>Then, they are obtained together. The embodiments of the present application do not limit this.
The merge sub-module 31 includes a merge sub-unit;
the merging subunit 31 is configured to merge the new case data and the samples of the original sample data set according to the original case data and the treatment effect label of each sample in the original sample data set, and the new case data and the predicted treatment effect.
When the merging subunit of the merging submodule 31 merges the new case data and the original sample data set, the features of the new case data and the features of the case data of a single sample in the original sample data set are learned and compared, or the features of the new case data, the predicted treatment effect, and the association between the features of the new case data and the predicted treatment effect are learned, and then the features of the original case data of a single sample in the original sample data set, the treatment effect, and the association between the features of the original case data and the treatment effect are learned, so as to complete the merging of the new case data and the original sample set.
For example, the characteristics of the new case data or the characteristics of the original case data may be the number of leukocytes in a blood routine laboratory sheet, brain CT data, runny nose, cough, penicillin, acyclovir, and the like. The predicted treatment effect is cure, and the treatment effect can be cure, semi-cure or no cure. The type of the treatment effect can be obtained by summarizing the possible treatment effects when the initial prediction model is trained, and the type of the treatment effect is set in the initial prediction model.
The original sample Data set, i.e. the initial case database DatadbThe merge with the new case data sample may be expressed as: dataset＝Datadb∪sample；
Wherein DatasetThe method refers to a sample set to be classified formed by merging an original sample data set (an initial case database) and new case data.
The clustering submodule 32 is configured to cluster the sample set to be classified to obtain a plurality of sample classifications;
the sample classification refers to different clusters obtained by segmenting case data in a sample set to be classified through a clustering algorithm.
The embodiment of the application can utilize a clustering algorithm based on division (such as k-means, k-models and the like), or a hierarchical clustering algorithm (such as CURE, ROCK and the like), or a density clustering algorithm (such as DBSCAN, GDBSCAN and the like), or a grid clustering algorithm (such as STING, WaveCluster and the like), or a neural network-based clustering algorithm (such as SOM (self-organizing neural network) and the like), or a statistical clustering algorithm (such as COOBE, C L ASSIT and the like), and the like, so as to cluster new case data and a large amount of original case data in a sample set to be classified, obtain N sample classifications, and enable the similarity of the case data in each sample classification to be as large as possible.
Exemplarily, the sample set Data to be classified is treated by a k-means clustering algorithmsetAfter clustering, 10 sample classifications (clusters) were obtained. Data clustered at this timesetEach case number in (a) has its corresponding sample classification.
The clustering sub-module 32 further comprises a labeling sub-unit for labeling all case data in the sample classification with the sample classification;
for the clustered DatasetAfter marking of the original case data in (1), each original case data can be expressed as<datai,clusteri>. When i is equal to 5, dataiRefers to the clustered DatasetThe 5 th case Data stored in (1), and the clustered DatasetThe 5 th case data ofiCorresponding to the sample class of clusteri。clusteriThe value of i in (1) is also 5. ClusteriThe corresponding sample classification may be any one of the 10 sample classifications. In other words, if the 5 th case dataiAt the 7 th sample classification, clusteriIf the 5 th case data is 7iAt sample class 10, clusteri＝10。
<datai,clusteri>Refers to the case dataiSample classified label case dataiFurthermore, for any case data, the sample classification where the case data is located can be directly found according to the mark. It will be appreciated that the same may be based on sample (new case data)And marking the sample, and directly determining the sample classification of the sample.
The clustering submodule 32 comprises a clustering subunit, which is configured to use the sample classification marked by the new case data as a first sample classification, and determine all case data included in the first sample classification as a target classification; wherein all case data comprised by the first sample classification belong to the set of samples to be classified.
Obtaining a sample classification clusterid corresponding to the sample marked with the sample classification, determining the sample classification of the sample by the clusterid, and for the sample classification of the sample, wherein each data isiMarker (2) clusteriAre all the same as the sample's label clusterid.
Suppose that the sample classification of sample is the 7 th sample classification among the 10 sample classifications mentioned above, and the label cluster of all case data in the 7 th sample classificationiAll with the new case data samplejAre labeled the same and are noted as: clusteri＝clusterid。
Classifying the first sample as DS;
DS＝{datai∣datai∈Datasetand clusteri＝clusterid}
the classification submodule 33 is configured to determine the sample classification in which the new case data is located as the target classification.
DS refers to a sample set Data to be classified formed by combining new case Data and original case DatasetAnd after clustering, classifying the samples of the new case data among the obtained multiple sample classifications.
DatasetThe method refers to a sample set to be classified formed by merging all original case data in an original sample data set (an initial case database) with new case data. Wherein set is a symbol referring to the sample set to be classified which distinguishes the original sample data set from the sample set to be classified.
dataiRefers to the original case data of any of the original sample data sets. Wherein i is the difference between different original cases in the original sample data setThe sign of the data.
clusteriThis means that any original case data is labeled with the sample classification to which it corresponds. Where i is a symbol that distinguishes different original case data in the original sample data set.
clusterid refers to the labeling of new case data with the sample classification to which it corresponds.
sample refers to new case data.
labelpredRefers to the predicted therapeutic effect of the new case data. Where pred is meant for prediction.
labeliRefers to the therapeutic effect of the original case data. Where i is a symbol that distinguishes different original case data in the original sample data set.
DatadbRefers to the original sample data. And db is a symbol which is used for distinguishing the original sample data set from the sample set to be classified and refers to the sample set to be original.
The meaning of DS is explained by way of example: according to the marker cluster of the sample, the sample is positioned in the 7 th cluster set, and the marker cluster of all case data in the DS is determinediAs with sample's marker clusterid, i.e., all case data in the DS is confirmed as samplejAlso in the 7 th cluster set, while confirming that all case Data in the DS are in the sample set Data to be classifiedsetIn (1).
The DS refers to a sample classification in which case data most similar to the new case data sample is located. And the labels of all case data in the first sample classification are the same as those of the new case data, and all case data in the sample classification are in the sample classification in the sample set to be classified. It can be understood that the first sample classification is not limited to a specific sample classification of a plurality of sample classifications obtained after clustering the sample set to be classified, for example, the first sample classification does not refer to the 1 st or 7 th sample classification or other sample classifications among the 10 sample classifications, but refers to the sample classification where the new case data sample is located.
Because all case data in each sample classification are most similar in a plurality of sample classifications obtained after clustering sample data to be classified, the case data of one sample classification is most similar to the new case data sample. Assuming that the new case data sample is classified in the 7 th sample, the case data in the 7 th sample classification is most similar to the new case data sample, and the similarity can be reflected in the similarity of the characteristics of the case data, such as the chief complaints are clear nasal discharge, red blood cells and white blood cells in blood routine are in a range, and the like. Therefore, the first sample classification is taken as a target classification, a set of case data with the most similar sample of the new case data is objectively determined, the target treatment effect of the sample of the new case data is obtained according to a large amount of case data with the most similar sample of the new case data, the experience of corresponding illness states of a plurality of case data is objectively summarized, and the scientificity of the treatment effect is ensured.
The case data classification module of the embodiment of the present application further provides a statistics submodule 34 and a labeling submodule 35 for obtaining a target treatment effect according to the target classification.
With continued reference to fig. 2, the statistics submodule 34 is operable to count the number of case data in the target category having the same therapeutic effect;
the labeling submodule 35 is configured to label the treatment effect with the largest number of case data as the target treatment effect.
The case data in the target classification may all have the same label of the treatment effect, or most of the case data may have the same label, and the labels of a small number of case data are different; for sample classification obtained by a clustering algorithm, a set of most similar case data is obtained, wherein a treatment effect label of single case data is the real treatment effect of a case or the treatment effect verified by medical staff, and the reliability is higher; meanwhile, as the treatment effect label of each case data is objective treatment effect (real treatment effect or treatment effect verified by medical staff), the treatment effects are different.
In view of this, the statistics submodule 34 is used to count the number of case data under each treatment effect label in the target classification, and then the treatment effect with the largest number of case data is used as the target treatment effect, so as to determine the treatment experience generally applicable to case data similar to the new case data.
Illustratively, assuming that the treatment effect labels of case data in the target classification are three types of "cured", "semi-cured" and "uncured", the number of case data having the "cured" label obtained statistically is N1The number of case data having the "semi-cure" label is N2The number of case data with the "cure" label is N3In which N is2>N3>N1And N is2>(N2+N3+N1) And/2, then the "half-cure" is the most targeted therapeutic effect.
The target treatment effect can also represent the classification sample of the new case data, i.e. the treatment effect of the target classification, so the target treatment effect is recorded as labelcluster. cluster is sample classification.
And after the case data classification module obtains the target treatment effect, replacing the predicted treatment effect of the new case data with the target treatment effect, and updating the label of the new case data.
Will be provided with<sample,labelpred>Is replaced by<sample,labelcluster>。
The new case data is taken as the data of the cases of nasosinusitis, the process of obtaining the target treatment effect 'semi-cure' is explained, the treatment effect prediction module predicts the new case data (the data of the cases of nasosinusitis) to obtain an initial label 'cure', and further the process of obtaining the target treatment effect 'semi-cure', and the treatment effect prediction module predicts the new case data (the data of the cases of nasosinusitis) to obtain the initial<Data on cases of sinusitis, cure>With the original sample data set<Raw case dataiTherapeutic effectsi>Merging, clustering a merged sample set to be classified containing new case data and original case data to obtain a plurality of sample classifications, taking the sample classification where the new case data is located as a target sample classification, taking the treatment effect with the largest number of case data in the target sample classification as a target treatment effect, and obtaining the target treatment effect of the new case data (sinusitis case data)The treatment effect is 'semi-cure', the target treatment effect is different from the predicted treatment effect at the moment, and the 'semi-cure' is taken as a new label of new case data and is recorded as<Data of cases of sinusitis, semi-cure>。
Meanwhile, the information that the target treatment effect is different from the predicted treatment effect triggers a prompting module of the treatment effect prediction system to generate a prompting message, and the prompting message is sent to an interactive page to prompt medical staff to verify the treatment effect of the new case data (sinusitis case data).
And the prompting module 8 is used for outputting a prompting message to prompt a user to verify the label of the new case data when the predicted treatment effect is different from the target treatment effect.
When labelpred≠labelclusterAnd then, the prompting module 8 outputs a corresponding prompting message to be sent to an interactive page to prompt a user to verify the label of the new case data.
The reminding message may be a pop-up window, a pop-up screen, or a voice message, and the like, which is not limited in the embodiment of the present application.
The system further comprises a verification module 7;
the checking module 7 is used for determining the label of the new case data according to the recorded actual treatment effect;
wherein the recorded therapeutic effect is the target therapeutic effect or the predicted therapeutic effect or a therapeutic effect other than the target therapeutic effect and the predicted therapeutic effect;
when the recorded treatment effect is other treatment effect, the database 5 takes the recorded treatment effect as the label of the new case data and updates the sample data set;
the treatment effect prediction module 2 is further configured to train and update the prediction model with the updated sample data set, and perform next treatment effect prediction by using the updated prediction model.
By way of example, assume that the current prediction model is a modelnowA therapeutic effect prediction module to include<samplej,labelcluster j>And<datai,labeli>sample set of (2), for modelnowUpdating to obtain modelnextThen with modelnextAs a prediction model for the therapeutic effect prediction module 2.
In the embodiment of the application, the set verification module enables medical staff to obtain detailed information of the new case data according to the reminding message, comprehensively considers the target treatment effect of the new case data and the rationality of predicting the treatment effect, and inputs the treatment effect.
The medical staff enters the treatment effect to select the predicted treatment effect or the target treatment effect, and can also input other treatment effects (the other treatment effects in the application refer to actual treatment results obtained by the medical staff and about the patient, so as to correct the target treatment effect or the predicted treatment effect obtained through the clustering algorithm in the database, and obtain a more accurate treatment effect prediction model).
If the treatment effect entered by the medical staff is the target treatment effect, the database 5 still takes the target treatment effect as the label of the new case data, and if the treatment effect entered by the medical staff is other treatment effect, the other treatment effect entered by the medical staff is taken as the label of the new case data.
It will be appreciated that the new case data stored by the database 5 preferably has a label of the effect of the treatment verified by the medical staff.
The database 5 is used for adding new case data with labels to obtain a sample data set;
the treatment effect prediction module 2 is further configured to train and update the prediction model with the sample data set, and predict a treatment effect of the next time by using the updated prediction model. Namely, the treatment effect prediction module in the application has the functions of self-detection and self-updating.
Illustratively, when the treatment effect prediction module detects that new case data newly added in the database or actual treatment results about the patient newly input by medical workers exist, the treatment effect prediction module automatically trains the treatment effect prediction model according to the newly changed data information to obtain an updated treatment effect prediction model, and performs next treatment effect prediction by using the updated prediction model.
Suppose that after medical staff obtains the prompting message that the predicted treatment effect 'cure' of the data of the cases of nasosinusitis is different from the target treatment effect 'semi-cure', the information of the cases of nasosinusitis is obtained, the treatment effect of the 'semi-cure' is obtained by checking, and the information is input into the checking module 7. The database 5 obtains the treatment effect returned by the check module 7, takes 'semi-cure' as a label, records the 'data of cases of nasosinusitis, semi-cure', and adds the 'data of cases of nasosinusitis, semi-cure' to the original sample data set.
And updating the original sample data set, triggering a treatment effect prediction module to add the sample data set of < data of cases of nasosinusitis and half-cure >, and training a prediction model in the module to obtain an updated prediction model. When the case data is input next time, the treatment effect prediction module 2 calculates the characteristics of the input case data with the updated prediction model to predict the treatment effect.
The embodiment of the application is provided with a case data classification module, the case data is automatically divided based on a clustering algorithm, and the target treatment effect of the new case data which is objectively input is obtained on the basis of a large amount of case data. The predicted treatment effect predicted by the system prediction model is detected according to the target treatment effect, and the prediction accuracy of the prediction model is corrected, so that the treatment effect prediction system has the capability of automatically detecting sample changes and automatically training according to the treatment effect obtained by automatically dividing case data. In addition, the embodiment of the application is also provided with a verification module to prompt medical staff to verify when the target treatment effect is different from the predicted treatment effect, so that the treatment effect prediction system can be automatically trained based on case division, has the capability of correcting errors based on sample data, and can be manually optimized in the process of using the prediction model. It can be understood that with continuous prediction and addition of new case data, the samples of the prediction model after multiple update iterations are more comprehensive, and the prediction model obtained by more comprehensive sample update is more accurate.
The embodiment of the present application further provides a therapeutic effect prediction system, and referring to fig. 3, fig. 3 is a therapeutic effect prediction system according to another embodiment of the present application.
A communication module 9, configured to receive a modification instruction for modifying a sample in the original sample data set;
except that when the treatment effect prediction module 2 predicts the actual treatment effect of the medical personnel entering the new case data, the case data classification module is triggered to classify all the case data in the database 5 so as to obtain the target treatment effect from the characteristic information of the case data, further test the accuracy of the prediction model and further optimize the model; medical personnel can also actively make modifications to the samples in the database 5 via the communication module 9.
The database 5 is further configured to update the original sample data set according to the modification instruction;
modification instructions include, but are not limited to, addition, subtraction, and modification of case data. Typical case data and actual treatment effect of other medical units can be collected as a new sample; the sample data set may be deleted, for example, the sample data set is cured at that time, but after a period of time, the patient relapses, and at this time, the corresponding case data in the sample data set needs to be modified, or the case data is directly deleted from the sample data set.
The treatment effect prediction module 2 is further configured to train and update the prediction model with the updated original sample data set when the original sample data set is updated, and perform next treatment effect prediction by using the updated prediction model.
In summary, the treatment effect prediction system provided by the embodiment of the application can predict the treatment effect according to the input case data, can also obtain the treatment effect which the case data should have compared with other case experiences based on the clustering algorithm, and further update the prediction model for obtaining the predicted treatment effect, so that the treatment effect prediction system has the error correction capability based on new case data samples, and continuously optimizes the prediction model per se; meanwhile, the treatment effect prediction system provided by the embodiment of the application is provided with the verification module, when the prediction treatment effect is inaccurate, the medical staff is reminded to verify the predicted treatment effect, and the prediction model is updated according to the verification result of the medical staff, so that the treatment effect prediction system provided by the embodiment of the application has the function of receiving feedback of the medical staff.
In addition, based on the treatment effect prediction system of the embodiment of the application, medical staff can actively modify case data in the system, and further enhance the sample error correction capability and the interaction performance of the treatment effect prediction system.
Based on the same concept, the embodiment of the present application further provides a therapeutic effect prediction terminal, where the therapeutic effect prediction terminal is an integrated terminal of part or all of the modules in the therapeutic effect prediction system in the embodiment of the first aspect;
referring to fig. 4, fig. 4 is a schematic structural diagram of a therapeutic effect prediction terminal according to an embodiment of the present application.
The treatment effect prediction terminal also comprises a display screen and an information input component, wherein the display screen is used for displaying the data output by the treatment effect prediction system; the information input component is used for a user to input data to the treatment effect prediction system.
The information input component may be a keyboard; the display screen and the information input component can also be a touch display screen with an integrated structure or a mobile computer.
The programs executed by the modules of the case data acquisition module 1, the treatment effect prediction module 2, the case data classification module 3, the treatment effect display module 4, the database 5 and the like included in the integrated terminal and the data used when the programs are executed can be stored by using a storage medium.
The modules of the case data acquisition module 1, the treatment effect prediction module 2, the case data classification module 3, the treatment effect display module 4, the database 5 and the like can be arranged in a server cluster formed by different processors and can also be arranged in the same processor.
The display screen may display information including, but not limited to: the medical data acquisition module receives new case data input by medical personnel; the treatment effect display module displays the predicted treatment effect output by the treatment effect prediction module; when the predicted treatment effect is different from the target treatment effect, the prompt module outputs a prompt message; the checking module records the actual treatment effect according to the reminding message received by the medical staff; the communication module receives a modification instruction indicating that the sample is to be modified.
The embodiments in the present specification are described in a progressive or descriptive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
While preferred embodiments of the present application have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the true scope of the embodiments of the application.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a device, system, or terminal apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such device, system, or terminal apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a device, system or terminal that comprises the element.
The treatment effect prediction system with the self-learning capability and the treatment effect prediction terminal provided by the application are introduced in detail, and the description of the embodiment is only used for helping to understand the core thought of the application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.
Claims (10)
1. A therapeutic effect prediction system, the system comprising: the system comprises a case data acquisition module, a treatment effect prediction module, a case data classification module, a treatment effect display module and a database; the case data acquisition module is connected with the treatment effect prediction module and the case data classification module;
the case data acquisition module is used for receiving and inputting new case data;
the treatment effect prediction module is used for obtaining a predicted treatment effect by using a prediction model according to the new case data and sending the predicted treatment effect to the treatment effect display module;
the case data classification module is used for obtaining a target treatment effect according to the target classification of the new case data and taking the target treatment effect as a label of the new case data;
the database is used for adding new case data with labels to obtain a sample data set;
the treatment effect prediction module is also used for training and updating the prediction model by the sample data set and predicting the next treatment effect by using the updated prediction model.
2. The treatment effect prediction system of claim 1, wherein the system further comprises:
a labeling module for labeling an initial label for the new case data, the initial label being the predicted treatment effect;
and the label updating module is used for replacing the predicted treatment effect with the target treatment effect and updating the label of the new case data when the predicted treatment effect is different from the target treatment effect.
3. The treatment effect prediction system of claim 2, characterized in that the system further comprises:
and the prompting module is used for outputting a prompting message to prompt a user to verify the label of the new case data when the predicted treatment effect is different from the target treatment effect.
4. The therapeutic effect prediction system of claim 3, wherein the system further comprises a verification module;
the checking module is used for determining the label of the new case data according to the recorded actual treatment effect;
wherein the recorded therapeutic effect is the target therapeutic effect or the predicted therapeutic effect or a therapeutic effect other than the target therapeutic effect and the predicted therapeutic effect;
when the recorded treatment effect is other treatment effect, the database takes the recorded treatment effect as the label of the new case data and updates the sample data set;
the treatment effect prediction module is also used for training and updating the prediction model by using the updated sample data set, and predicting the treatment effect of the next time by using the updated prediction model.
5. The treatment effect prediction system of claim 1,
the database stores an original sample data set;
the case data classification module comprises a merging submodule, a clustering submodule and a classification submodule;
the merging submodule is used for acquiring the original sample data set from the database and merging the new case data and the original sample data set to obtain a sample set to be classified;
the clustering submodule is used for clustering the sample set to be classified to obtain a plurality of sample classifications;
the classification submodule is used for determining the sample classification where the new case data is located as the target classification.
6. The treatment effect prediction system of claim 5,
a single sample in the original sample data set is original case data carrying a treatment effect label;
the merging submodule comprises a merging subunit;
the merging subunit is configured to merge the new case data and the samples of the original sample data set according to the original case data and the treatment effect label of each sample in the original sample data set, and the new case data and the predicted treatment effect.
7. The treatment effect prediction system of claim 6, wherein the case data classification module further comprises a statistics submodule and a labeling submodule;
the statistic submodule is used for counting the number of case data with the same treatment effect in the target classification;
the labeling submodule is configured to label the treatment effect with the largest number of case data as the target treatment effect.
8. The treatment effect prediction system of claim 6,
the clustering sub-module further comprises a labeling sub-unit for labeling all case data in the sample classification with the sample classification;
the clustering submodule comprises a clustering subunit, and the clustering subunit is used for taking the sample classification marked by the new case data as a first sample classification and determining all case data included in the first sample classification as a target classification; wherein all case data comprised by the first sample classification belong to the set of samples to be classified.
9. The treatment effect prediction system of claim 5, characterized in that the system further comprises:
the communication module is used for receiving a modification instruction for modifying the samples in the original sample data set;
the database is also used for updating the original sample data set according to the modification instruction;
the treatment effect prediction module is also used for training and updating the prediction model by the updated original sample data set when the original sample data set is updated, and predicting the treatment effect of the next time by using the updated prediction model.
10. A therapeutic effect prediction terminal characterized in that the therapeutic effect prediction terminal is an integrated terminal including part or all of the modules in the therapeutic effect prediction system according to any one of claims 1 to 9;
the treatment effect prediction terminal also comprises a display screen and an information input component, wherein the display screen is used for displaying the data output by the treatment effect prediction system; the information input component is used for a user to input data to the treatment effect prediction system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010167791.2A CN111477321B (en) | 2020-03-11 | 2020-03-11 | Treatment effect prediction system with self-learning capability and treatment effect prediction terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010167791.2A CN111477321B (en) | 2020-03-11 | 2020-03-11 | Treatment effect prediction system with self-learning capability and treatment effect prediction terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111477321A true CN111477321A (en) | 2020-07-31 |
| CN111477321B CN111477321B (en) | 2023-06-09 |
Family
ID=71747327
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010167791.2A Active CN111477321B (en) | 2020-03-11 | 2020-03-11 | Treatment effect prediction system with self-learning capability and treatment effect prediction terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111477321B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116798646A (en) * | 2023-08-17 | 2023-09-22 | 四川互慧软件有限公司 | Snake injury prognosis prediction method and device based on clustering algorithm and electronic equipment |
| CN117916816A (en) * | 2022-08-19 | 2024-04-19 | 优特心研究所股份有限公司 | Computer heart disease database operating method, computer heart disease database operating program, and information processing device |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1760881A (en) * | 2005-11-14 | 2006-04-19 | 南京大学 | Modeling method of forecast in device of computer aided diagnosis through using not diagnosed cases |
| KR20140090483A (en) * | 2013-01-09 | 2014-07-17 | 경희대학교 산학협력단 | Method for clustering health-information |
| US20140297642A1 (en) * | 2009-02-10 | 2014-10-02 | Ayasdi, Inc. | Systems and methods for mapping patient data from mobile devices for treatment assistance |
| CN104915560A (en) * | 2015-06-11 | 2015-09-16 | 万达信息股份有限公司 | Method for disease diagnosis and treatment scheme based on generalized neural network clustering |
| CN106294343A (en) * | 2015-05-13 | 2017-01-04 | 富士通株式会社 | Data clustering method, model fusion method and device |
| RU2609737C1 (en) * | 2015-12-21 | 2017-02-02 | Общество с ограниченной ответственностью "СИАМС" | Automated system of distributed cognitive support of making diagnostic decisions in medicine |
| CN107358019A (en) * | 2017-05-25 | 2017-11-17 | 上海交通大学医学院附属瑞金医院 | Medical scheme recommendation system and method suitable for concept drift |
| CN107895168A (en) * | 2017-10-13 | 2018-04-10 | 平安科技（深圳）有限公司 | The method of data processing, the device of data processing and computer-readable recording medium |
| CN108520780A (en) * | 2018-03-07 | 2018-09-11 | 中国科学院计算技术研究所 | A kind of medical data processing and system based on transfer learning |
| WO2019136806A1 (en) * | 2018-01-12 | 2019-07-18 | 平安科技（深圳）有限公司 | Medical model training method and apparatus, medical identification method and apparatus, device, and medium |
| US20190237200A1 (en) * | 2016-10-18 | 2019-08-01 | Fujitsu Limited | Recording medium recording similar case retrieval program, information processing apparatus, and similar case retrieval method |
| CN110111888A (en) * | 2019-05-16 | 2019-08-09 | 闻康集团股份有限公司 | A kind of XGBoost disease probability forecasting method, system and storage medium |
| CN110807159A (en) * | 2019-10-30 | 2020-02-18 | 同盾控股有限公司 | Data marking method and device, storage medium and electronic equipment |
-
2020
- 2020-03-11 CN CN202010167791.2A patent/CN111477321B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1760881A (en) * | 2005-11-14 | 2006-04-19 | 南京大学 | Modeling method of forecast in device of computer aided diagnosis through using not diagnosed cases |
| US20140297642A1 (en) * | 2009-02-10 | 2014-10-02 | Ayasdi, Inc. | Systems and methods for mapping patient data from mobile devices for treatment assistance |
| KR20140090483A (en) * | 2013-01-09 | 2014-07-17 | 경희대학교 산학협력단 | Method for clustering health-information |
| CN106294343A (en) * | 2015-05-13 | 2017-01-04 | 富士通株式会社 | Data clustering method, model fusion method and device |
| CN104915560A (en) * | 2015-06-11 | 2015-09-16 | 万达信息股份有限公司 | Method for disease diagnosis and treatment scheme based on generalized neural network clustering |
| RU2609737C1 (en) * | 2015-12-21 | 2017-02-02 | Общество с ограниченной ответственностью "СИАМС" | Automated system of distributed cognitive support of making diagnostic decisions in medicine |
| US20190237200A1 (en) * | 2016-10-18 | 2019-08-01 | Fujitsu Limited | Recording medium recording similar case retrieval program, information processing apparatus, and similar case retrieval method |
| CN107358019A (en) * | 2017-05-25 | 2017-11-17 | 上海交通大学医学院附属瑞金医院 | Medical scheme recommendation system and method suitable for concept drift |
| CN107895168A (en) * | 2017-10-13 | 2018-04-10 | 平安科技（深圳）有限公司 | The method of data processing, the device of data processing and computer-readable recording medium |
| WO2019136806A1 (en) * | 2018-01-12 | 2019-07-18 | 平安科技（深圳）有限公司 | Medical model training method and apparatus, medical identification method and apparatus, device, and medium |
| CN108520780A (en) * | 2018-03-07 | 2018-09-11 | 中国科学院计算技术研究所 | A kind of medical data processing and system based on transfer learning |
| CN110111888A (en) * | 2019-05-16 | 2019-08-09 | 闻康集团股份有限公司 | A kind of XGBoost disease probability forecasting method, system and storage medium |
| CN110807159A (en) * | 2019-10-30 | 2020-02-18 | 同盾控股有限公司 | Data marking method and device, storage medium and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 马光志，等: "基于聚类和分类的自学习系统模型", pages 83 - 84 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117916816A (en) * | 2022-08-19 | 2024-04-19 | 优特心研究所股份有限公司 | Computer heart disease database operating method, computer heart disease database operating program, and information processing device |
| CN116798646A (en) * | 2023-08-17 | 2023-09-22 | 四川互慧软件有限公司 | Snake injury prognosis prediction method and device based on clustering algorithm and electronic equipment |
| CN116798646B (en) * | 2023-08-17 | 2023-11-24 | 四川互慧软件有限公司 | Snake injury prognosis prediction method and device based on clustering algorithm and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111477321B (en) | 2023-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11790171B2 (en) | Computer-implemented natural language understanding of medical reports | |
| US11081215B2 (en) | Medical record problem list generation | |
| WO2020214683A1 (en) | Computer-implemented natural language understanding of medical reports | |
| US8612261B1 (en) | Automated learning for medical data processing system | |
| EP3977343A1 (en) | Systems and methods of clinical trial evaluation | |
| US20200356846A1 (en) | Prediction of healthcare outcomes and recommendation of interventions using deep learning | |
| CN105009126A (en) | Medical database and system | |
| US20180210925A1 (en) | Reliability measurement in data analysis of altered data sets | |
| US20230005574A1 (en) | Methods and systems for comprehensive symptom analysis | |
| US20150149215A1 (en) | System and method to detect and visualize finding-specific suggestions and pertinent patient information in radiology workflow | |
| US11923075B2 (en) | System and method associated with determining physician attribution related to in-patient care using prediction-based analysis | |
| CN111477321B (en) | Treatment effect prediction system with self-learning capability and treatment effect prediction terminal | |
| CN111477320A (en) | Construction system of treatment effect prediction model, treatment effect prediction system and terminal | |
| KR102506643B1 (en) | Patient information based medication management device and method | |
| CN110471941B (en) | Method and device for automatically positioning judgment basis and electronic equipment | |
| Ben-Assuli et al. | Data mining techniques utilizing latent class models to evaluate emergency department revisits | |
| WO2023217737A1 (en) | Health data enrichment for improved medical diagnostics | |
| CN116825311A (en) | DRG/DIP-based hospital management and control operation method and system | |
| CN111400759A (en) | Visiting time table generation method and device, storage medium and electronic equipment | |
| Na et al. | Patient outcome predictions improve operations at a large hospital network | |
| CN112530585A (en) | Data processing method and device based on medical institution, computer equipment and medium | |
| US20200126641A1 (en) | Medical concierge | |
| CN115151182B (en) | Method and system for diagnostic analysis | |
| CN111063436A (en) | Data processing method and device, storage medium and electronic terminal | |
| CN110223771A (en) | Gastroenterology electronic data analysis method based on NHPP |
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 |