CN112071431A - Clinical path automatic generation method and system based on deep learning and knowledge graph - Google Patents
Clinical path automatic generation method and system based on deep learning and knowledge graph Download PDFInfo
- Publication number
- CN112071431A CN112071431A CN202010989064.4A CN202010989064A CN112071431A CN 112071431 A CN112071431 A CN 112071431A CN 202010989064 A CN202010989064 A CN 202010989064A CN 112071431 A CN112071431 A CN 112071431A
- Authority
- CN
- China
- Prior art keywords
- medical
- intersection
- deep learning
- clinical
- clinical path
- 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
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000013135 deep learning Methods 0.000 title claims abstract description 31
- 239000003814 drug Substances 0.000 claims abstract description 47
- 238000013136 deep learning model Methods 0.000 claims abstract description 46
- 229940079593 drug Drugs 0.000 claims abstract description 21
- 238000009826 distribution Methods 0.000 claims abstract description 14
- 239000013598 vector Substances 0.000 claims description 61
- 230000037361 pathway Effects 0.000 claims description 35
- 201000010099 disease Diseases 0.000 claims description 32
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 32
- 238000003745 diagnosis Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 10
- 238000013507 mapping Methods 0.000 claims description 8
- 230000002457 bidirectional effect Effects 0.000 claims description 7
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000007477 logistic regression Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000012417 linear regression Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 238000011176 pooling Methods 0.000 claims 2
- 230000008569 process Effects 0.000 description 12
- 238000012549 training Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 11
- 238000012795 verification Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 238000011272 standard treatment Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 239000000203 mixture Substances 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 239000000306 component Substances 0.000 description 4
- 230000007170 pathology Effects 0.000 description 3
- 239000005426 pharmaceutical component Substances 0.000 description 3
- 230000000306 recurrent effect Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 210000004072 lung Anatomy 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000002271 resection Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 208000011580 syndromic disease Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229940044683 chemotherapy drug Drugs 0.000 description 1
- 210000001072 colon Anatomy 0.000 description 1
- 229940126534 drug product Drugs 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- QOFFJEBXNKRSPX-ZDUSSCGKSA-N pemetrexed Chemical compound C1=N[C]2NC(N)=NC(=O)C2=C1CCC1=CC=C(C(=O)N[C@@H](CCC(O)=O)C(O)=O)C=C1 QOFFJEBXNKRSPX-ZDUSSCGKSA-N 0.000 description 1
- 229960005079 pemetrexed Drugs 0.000 description 1
- 239000000825 pharmaceutical preparation Substances 0.000 description 1
- 238000011301 standard therapy Methods 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
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
- G16H50/70—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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Pathology (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Databases & Information Systems (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention provides a clinical path automatic generation method and system based on deep learning and a knowledge graph. The method comprises the steps of inputting patient characteristic information into a first deep learning model and a second deep learning model respectively, and recommending a first clinical path and a second clinical path correspondingly; taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set; taking an intersection of the medical operation advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; and obtaining the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
Description
Technical Field
The invention belongs to the field of clinical data processing, and particularly relates to a clinical path automatic generation method and system based on deep learning and knowledge maps.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
At present, two methods for generating clinical paths are available, one is to directly use the clinical paths of various diseases of national standards, and the other is to simply analyze the existing diagnosis and treatment schemes of doctors to make the clinical paths of various diseases by using the traditional machine learning method.
The inventor finds that the clinical path uniformly established according to the national standard does not combine the actual conditions of each hospital, department, doctor and patient, and cannot meet the various conditions of the patient, so that the clinical path cannot meet the requirement of actual treatment, and therefore, the clinical path is difficult to play a role in actual clinical work. The clinical path established based on the traditional machine learning method is not accurate enough due to the fact that the analysis range is not wide enough and the model complexity is not enough, and cannot meet the requirements of clinical use of doctors.
Disclosure of Invention
In order to solve the problems, the invention provides a clinical pathway automatic generation method and system based on deep learning and a knowledge graph, which are based on deep learning and a knowledge graph and on standardized treatment data, intelligently recommend clinical pathways suitable for patients of the type, and can improve the accuracy of the formulated clinical pathways.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a clinical path automatic generation method based on deep learning and knowledge mapping.
In one or more embodiments, a method for automatic generation of clinical pathways based on deep learning and knowledge-maps includes:
respectively inputting the patient characteristic information into a first deep learning model and a second deep learning model, and correspondingly recommending a first clinical path and a second clinical path; wherein the first clinical path only contains the medical order, and the second clinical path contains the medical order and the execution time thereof;
taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set;
taking an intersection of the medical operation advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information;
and obtaining the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
A second aspect of the invention provides an automated clinical pathway generation system based on deep learning and knowledge mapping.
In one or more embodiments, a system for automated generation of clinical pathways based on deep learning and knowledge-maps, comprising:
the clinical path recommending module is used for inputting the patient characteristic information into the first deep learning model and the second deep learning model respectively and recommending a first clinical path and a second clinical path correspondingly; wherein the first clinical path only contains the medical order, and the second clinical path contains the medical order and the execution time thereof;
the first intersection taking module is used for taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set;
the second intersection taking module is used for taking an intersection of the medical operation medical advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information;
and the medical advice distribution module is used for acquiring the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
A third aspect of the invention provides a computer-readable storage medium.
A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method for automatic generation of a clinical pathway based on deep learning and knowledge-maps as set forth above.
A fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the steps in the method for automatic generation of a clinical pathway based on deep learning and knowledge maps as described above.
Compared with the prior art, the invention has the beneficial effects that:
the patient characteristic information is respectively input into a first deep learning model and a second deep learning model, and a first clinical path and a second clinical path are correspondingly recommended; the first clinical path only contains the medical advice, the second clinical path contains the medical advice and the execution time of the medical advice, patient characteristic information provided by a doctor is comprehensively analyzed based on a deep learning technology, and clinical paths suitable for the patients are intelligently recommended based on standardized treatment data, so that the accuracy of the formulated clinical paths is greatly improved;
the method is also based on a standard clinical path corresponding to the patient disease diagnosis name in the knowledge map and the patient characteristic information, and the medical operation alias set and the medicine commodity name set respectively correspond to the medical operation medical advice recommendation set and the medicine commodity name set to obtain an intersection set, so that a first intersection set and a second intersection set are obtained; the method comprises the steps of obtaining execution time of each medical advice in a first intersection and a second intersection through a second clinical path, obtaining medical advice distribution of the first intersection and the second intersection corresponding to the clinical path every day/every stage, automatically recommending a treatment scheme commonly used by a doctor for reference when the doctor formulates the clinical path, formulating the whole treatment process aiming at the clinical path, and providing a feasible clinical path for the doctor based on historical standard treatment data of the hospital.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a flow chart of a method for automatic generation of clinical pathways based on deep learning and knowledge maps in accordance with an embodiment of the present invention;
FIG. 2 is a diagram of a first deep learning model according to an embodiment of the invention;
FIG. 3 is a diagram of a second deep learning model according to an embodiment of the present invention;
FIG. 4 is a structural diagram of an automatic clinical pathway generation system based on deep learning and knowledge mapping according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Interpretation of terms:
structural characteristics: for a certain feature of the patient, if the feature only needs structured data for expression, the feature is called as a structured feature, and the medical advice information is structured feature information;
unstructured features: for a certain feature of a patient, if unstructured text data is needed for expressing the feature, the feature is called an unstructured feature, and description information of a disease symptom in the current medical history is unstructured feature information.
Medical operation standard name and medical operation alias: the medical operation type order has both standard and alias names, for example, "left superior lung lobe resection" is the standard name of a certain operation, "left superior lung lobe resection" is the alias name of the operation, and one order has only one standard name and can have a plurality of alias names.
Pharmaceutical ingredient name and drug product name: for drugs of different compositions, the names of the compositions are recorded in the national standard clinical pathway, but in the clinical work of doctors, the name of the prescribed drug is often the name of the product of the drug rather than the name of the composition, for example, "oxitinib" is the name of a drug composition, and "tirisate" is the name of the product of the composition.
Knowledge graph S: the data related to the unstructured characteristic information refers to a text part describing the characteristic, the texts contain keywords and words related to the index, and the knowledge graph S records the keywords and words of different unstructured characteristic information.
Knowledge graph T: name of disease diagnosis- > standard clinical pathway, recording all orders per day or stage in the standard clinical pathway uniformly established by its country for each disease.
Knowledge graph Y: the pharmaceutical component name- > pharmaceutical commodity name records different pharmaceutical commodity names corresponding to different pharmaceutical component names, and each pharmaceutical component name can contain a plurality of pharmaceutical commodity names.
Knowledge graph C: medical operation standard name- > medical operation alias, and different alias names corresponding to the standard names of different medical operation orders are recorded according to the standard names.
Example one
Referring to fig. 1, the method for automatically generating a clinical pathway based on deep learning and a knowledge graph of the present embodiment includes:
step S101: respectively inputting the patient characteristic information into a first deep learning model and a second deep learning model, and correspondingly recommending a first clinical path and a second clinical path; wherein the first clinical pathway includes only the medical order and the second clinical pathway includes the medical order and its execution time.
In the specific implementation, patient characteristic information provided by a doctor is received, the information is recorded as a set A, and a disease diagnosis name in the patient characteristic is recorded as D.
The feature set A is input into a first deep learning model M and a second deep learning model G, respectively, and the two models output recommended clinical paths R _ M and R _ G, respectively. Wherein, the two clinical paths R _ M and R _ G are the full-flow treatment scheme, wherein R _ G includes the execution time of each order.
In a specific implementation, the first deep learning model M is used for receiving various types of feature information of a patient and deriving a clinical pathway suitable for the patient, i.e. a full-flow treatment scheme.
As shown in fig. 2, the first deep learning model M includes:
the bidirectional long-time memory cyclic neural network model is used for receiving unstructured text information word vectors of the characteristic information of the patient and enabling output results to pass through a maximum pool to obtain disease description vectors representing the patient;
the logistic regression model is used for combining the vector of the patient characteristic information structured information and the patient disease description vector, and mapping the combined vector to an index vector representing the similarity relation between the information of the patient and all medical orders; wherein any element in the exponential vector represents a probability of the corresponding order entering the clinical pathway.
Specifically, the unstructured text information word vector (possibly a plurality of) of the patient p is input into a bidirectional long-time memory recurrent neural network model, and the output results of all output word vectors are processed by a maximum pool technology to obtain a vector C \ucapable of representing the disease condition description of the patient pp(ii) a Vector L _, of structured information of patient ppAnd C \ upCombined to define a vector V _p＝[L_p,C_p]。
Vector V \uis transformed by using logistic regression methodpMapping to an exponential vector p representing the similarity between the information of the patient p and all orders, i.e.
Where c1, c2 are the parameter tensors in the logistic regression relationship that need to be trained to determine specific numerical values, the dimension of ρ is the number of all possible orders, and the ith element ρ (i) in the vector ρ represents the probability of selecting the ith order into the clinical path. The cross entropy of ρ and the 0-1 probability distribution q of all order results is taken as the loss function of the model M.
Before the training of the first deep learning model M, information vectors corresponding to 60%, 20% and 20% of k patients corresponding to each disease in the standard treatment information base H are respectively classified into a training set, a verification set and a test set. And training the first deep learning model M by utilizing the training set, the verification set and the test set.
The whole training process of the first deep learning model M is as follows: storing the obtained model once per preset training times (such as 100 times) on a training set; after the primary model is saved, the model is used for primary verification on the verification set to obtain and save the sum of loss functions of all patients in the verification set; after verifying a preset number of times (for example, 100 times), selecting a parameter corresponding to the model with the minimum verification loss function sum as a final parameter of the first deep learning model M.
The construction process of the first deep learning model M is as follows:
step (a 1): and constructing a standard treatment information base H of the hospital, wherein the standard treatment information base H comprises years of historical diagnosis and treatment data of the hospital, and the data comprises information such as basic information, medical records, first pages of medical records, medical advice, examination reports, inspection reports, pathology reports, electrocardiogram reports and the like of different patients.
Specifically, the step (a1) is implemented by the following steps:
for each disease diagnosis name in the knowledge map T, k (for example, k is 100) different patients diagnosed with the disease are found in the hospital historical clinical information base, and information such as basic information, medical history, medical record top page, medical advice, examination report, pathology report, electrocardiogram report, and the like of the patients is stored in the information base H. The information base H has a storage structure of disease diagnosis names- > patient numbers (k patients for each syndrome name) > basic information of a patient with a certain number, medical records, medical record first pages, medical advice, examination reports, pathology reports, electrocardiogram reports, and the like.
Wherein: the definition of "different patients" is that the similarity calculated using the Levenshtein distance between the admission record texts of each patient is less than a certain threshold, such as 0.5.
For the patient p, the information mainly includes two types, one is structured information (including the sex, age, syndrome, and name of disease diagnosis of the patient), and the other is unstructured text information (including medical record, examination report, etc. of the patient).
For the structured information, the structured information is converted into vector information (called a category vector) according to a 0-1 vector corresponding to the category, that is, the structured information can be represented by a certain vector, for example, as follows:
example 1: the sex of the patient is totally two types (male and female), the male category vector is (1,0), and the female category vector is (0,1), so that the sex information of the patient is converted from "male or female" to vector (1,0) or (0,1) ".
Example 2: assuming that the information base H has 3000 disease diagnosis names, the category vectors corresponding to the disease diagnosis names of the patients have 3000 dimensions, the dimension values corresponding to the disease diagnoses of the patients are 1, and the remaining dimension values are 0.
For patient p, all structured information is converted into a category vector, and all category vectors of the patient are combined into one vector, which is marked as L \_p。
It should be noted that, in other embodiments, the first deep learning model M may also be implemented in other existing structural forms, and will not be described here again.
For unstructured text information, because the text information is large in quantity and contains a large amount of irrelevant information, important relevant information in the text information is firstly screened and extracted (the method for extracting the information is shown in step iii below), and then vectorization representation of the extracted text information is completed, namely words in disease description are represented by word vectors by word segmentation technologies (such as Chinese segmentation), word vector technologies (such as word2vec and BERT).
The unstructured text information with a certain characteristic f is divided according to colons, commas, semicolons and periods, and each small segment is called a small sentence. For each clause, it is input into the third deep learning model K _ f. And (4) a corresponding third deep learning model K _ f is arranged for each unstructured feature f, and whether the small sentence is extracted or not is judged. And extracting all the small sentences related to the features f, splicing the small sentences into a text, and outputting the text.
Wherein, the structure of the third deep learning model K _ f is:
the word vector set of a certain sentence is input into a bidirectional gated loop network (Bi-GRU), the output results of all output word vectors are subjected to a 'maximum pool technique' to obtain a vector M _ f, and the vector M _ f is subjected to linear regression and is mapped to a probability p representing whether the sentence is extracted, namely p is c 1M _ f + c2, wherein c1 and c2 are parameter tensors in the linear regression relationship and need to be trained to determine specific numerical values. The cross entropy of p and the true 0-1 probability q of whether the clause is extracted is taken as the loss function of the model K _ f.
The application process of the third deep learning model K _ f is as follows:
acquiring a standard treatment information base H of a hospital, inputting a certain small sentence into a deep learning model K _ f, outputting the probability p of the small sentence being extracted, if the probability p is greater than a preset threshold (for example, 0.5, other values can be used), judging that the small sentence is extracted, and otherwise, judging that the small sentence is not extracted.
Specifically, the process of training the third deep learning model K _ f is as follows:
dividing word vector sets corresponding to 60%, 20% and 20% of all the sentences related to the characteristics f of k patients corresponding to each disease in the standard treatment information base H into a training set, a verification set and a test set respectively;
the whole training process is as follows: storing the obtained model once per preset training times (such as 100 times) on a training set; after the primary model is saved, the model is used for primary verification on the verification set to obtain and save the sum of loss functions of all patients in the verification set; after verifying for a preset number of times (for example, 100 times), selecting a parameter corresponding to the model with the minimum verification loss function sum as a final parameter of the third deep learning model K _ f.
It should be noted here that, in other embodiments, the third deep learning model K _ f may also be implemented in other existing structural forms, and will not be described here again.
Step (a 2): and inputting the patient characteristic information given by the doctor into the deep learning model M, and outputting the probability vectors rho of all the medical orders.
Step (a 3): and selecting the order corresponding to the dimension of the vector rho with the probability larger than the preset probability value (such as 0.5), otherwise, not selecting the order. And outputting all the selected orders.
As shown in fig. 3, the second deep learning model G is used to receive various feature information of the patient, derive a clinical path suitable for the patient and an execution time of each order, i.e., a full-flow treatment plan.
The second deep learning model G is an optimization model based on the first deep learning model M, and the construction process of the second deep learning model G is as follows:
step (b 1): constructing a standard treatment information base H of the hospital;
step (b 2): for each patient p in the standard therapy information base H, its information is entered into a trained model M, the vector V _uof whichpAnd establishing a vector library V corresponding to the information library H.
Specifically, the unstructured text information word vector (possibly a plurality of) of the patient p is input into a bidirectional long-time memory recurrent neural network model, and the output results of all output word vectors are processed by a maximum pool technology to obtain a vector C \ucapable of representing the disease condition description of the patient pp(ii) a Vector L _, of structured information of patient ppAnd C \ upCombined to define a vector V _p＝[L_p,C_p]。
Step (b 3): inputting the characteristic information of the patient u input by the doctor into the trained model M to obtain a vector V \uu。
Inputting the (possibly a plurality of) unstructured text information word vectors of the patient u into a bidirectional long-time memory recurrent neural network model, and obtaining a vector C \uwhich can represent the disease description of the patient u by passing the output results of all output word vectors through a maximum pool technologyu(ii) a Vector L \uof structured information of patient uuAnd C \ uuCombined to define a vector V _u＝[L_u,C_u]。
The following optimization problems are solved:
and taking all medical orders of the patient p corresponding to the optimal solution V _ p of the optimization problem in the information base H and the execution time of each medical order as the output of the model G.
Step S102: and taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to the categories, and forming a medical operation medical order recommendation set and a medicine medical order recommendation set.
Specifically, the intersection of all the medical orders contained in R _ M, R _ G is used to obtain a set B, where the medical orders in the set B are classified into two types, one type is medical procedure orders (e.g., surgery, note that the name of the medical procedure order here may be an alias name and not necessarily a standard name), and the other type is medicine orders (e.g., chemotherapy drug pemetrexed, note that the name of the medicine order here is a commodity name of the medicine and not necessarily a medicine component name), and the sets of the medical procedure and the medicine orders in the set B are respectively denoted as B _1 and B _ 2.
Step S103: taking an intersection of the medical operation advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information.
Specifically, in a knowledge graph T, a standard clinical path corresponding to the disease diagnosis name D of the patient in the step 1 is searched, and a set consisting of medical operations and medical orders of medicines in the standard clinical path is respectively recorded as E _1_ o and E _2_ o; here, the medical operation name included in E _1_ o refers to a medical operation standard name, and the medicine included in E _2_ o refers to a medicine component name rather than a medicine commodity name.
For the medical operation standard names contained in E _1_ o, corresponding medical operation aliases are searched in the knowledge graph C, and the set formed by all the found medical operation aliases is recorded as E _ 1.
For the medicine components contained in E _2_ o, corresponding medicine commodity names are searched in the knowledge graph Y, and the set formed by all the commodity names is marked as E _ 2.
Taking intersection of B _1 and E _1 to obtain a set C _ 1; and taking intersection of the B _2 and the E _2 to obtain a set C _ 2.
Step S104: and obtaining the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
Specifically, for each order in C _1 and C _2, the execution time is obtained through R _ G, and the order distribution of each day/stage of the clinical pathway C _1 and C _2 is obtained according to the time.
The patient characteristic information provided by a doctor is comprehensively analyzed based on the deep learning technology, the clinical path suitable for the patient is intelligently recommended based on the standardized treatment data, and the accuracy of the formulated clinical path is greatly improved; searching for a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information, and respectively corresponding to the medical operation alias set and the medicine commodity name set to obtain an intersection set to obtain a first intersection set and a second intersection set; the method comprises the steps of obtaining execution time of each medical advice in a first intersection and a second intersection through a second clinical path, obtaining medical advice distribution of the first intersection and the second intersection corresponding to the clinical path every day/every stage, automatically recommending a treatment scheme commonly used by a doctor for reference when the doctor formulates the clinical path, formulating the whole treatment process aiming at the clinical path, and providing a feasible clinical path for the doctor based on historical standard treatment data of the hospital.
Example two
Referring to fig. 4, the present embodiment provides an automatic clinical pathway generation system based on deep learning and knowledge-graph, which includes:
(1) the clinical path recommending module is used for recommending the patient characteristic clinical path, inputting the patient characteristic information into the first deep learning model and the second deep learning model respectively, and correspondingly recommending a first clinical path and a second clinical path; wherein the first clinical path only contains the medical order, and the second clinical path contains the medical order and the execution time thereof;
(2) the first intersection taking module is used for taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set;
(3) the second intersection taking module is used for taking an intersection of the medical operation medical advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information;
(4) and the medical advice distribution module is used for acquiring the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
The system for automatically generating clinical pathway based on deep learning and knowledge graph of this embodiment corresponds to the steps of the method for automatically generating clinical pathway based on deep learning and knowledge graph of the first embodiment one by one, and the specific implementation process is as described in the first embodiment, and will not be described here again.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the method for automatic generation of clinical pathway based on deep learning and knowledge-maps as described in the first embodiment above.
Example four
The embodiment provides a computer device, which comprises a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor executes the program to realize the steps of the method for automatically generating the clinical path based on the deep learning and knowledge mapping according to the embodiment.
As will be appreciated by one skilled in the art, 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 a hardware embodiment, a 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, 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 flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams 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.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A clinical pathway automatic generation method based on deep learning and knowledge graph is characterized by comprising the following steps:
respectively inputting the patient characteristic information into a first deep learning model and a second deep learning model, and correspondingly recommending a first clinical path and a second clinical path; wherein the first clinical path only contains the medical order, and the second clinical path contains the medical order and the execution time thereof;
taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set;
taking an intersection of the medical operation advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information;
and obtaining the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
2. The method of claim 1, wherein the medical procedure name included in the standard clinical pathway is a medical procedure standard name, and the drug name included in the standard clinical pathway is a drug component name.
3. The method for automatic generation of clinical pathways based on deep learning and knowledge-maps according to claim 1, wherein the patient characteristic information comprises structured information and unstructured textual information.
4. The method for automatic generation of clinical pathways based on deep learning and knowledge-graphs as claimed in claim 1, wherein the first deep learning model comprises:
the bidirectional long-time memory cyclic neural network model is used for receiving unstructured text information word vectors of the characteristic information of the patient and enabling output results to pass through a maximum pool to obtain disease description vectors representing the patient;
the logistic regression model is used for combining the vector of the patient characteristic information structured information and the patient disease description vector, and mapping the combined vector to an index vector representing the similarity relation between the information of the patient and all medical orders; wherein any element in the exponential vector represents a probability of the corresponding order entering the clinical pathway.
5. The method for automatic generation of clinical pathway based on deep learning and knowledge-graph of claim 4 wherein unstructured text information word vectors of patient feature information are extracted by a third deep learning model.
6. The method for automatic generation of clinical pathways based on deep learning and knowledge-graphs as claimed in claim 5, wherein the third deep learning model comprises:
the bidirectional gating circulation network is used for receiving a word vector set of a certain clause and enabling output results of all output word vectors to pass through a maximum pool to obtain a maximum pooling vector;
a linear regression model for linearly mapping the largest pooling vector to a probability representing whether the clause was extracted or not extracted.
7. The method for automatic generation of clinical pathway based on deep learning and knowledge-graph of claim 1, characterized in that the second deep learning model is an optimization model based on the first deep learning model.
8. An automatic generation system of clinical pathway based on deep learning and knowledge graph, comprising:
the clinical path recommending module is used for inputting the patient characteristic information into the first deep learning model and the second deep learning model respectively and recommending a first clinical path and a second clinical path correspondingly; wherein the first clinical path only contains the medical order, and the second clinical path contains the medical order and the execution time thereof;
the first intersection taking module is used for taking an intersection of all medical orders contained in the first clinical path and the second clinical path, dividing the medical orders in the intersection according to categories and forming a medical operation medical order recommendation set and a medicine medical order recommendation set;
the second intersection taking module is used for taking an intersection of the medical operation medical advice recommendation set and the medical operation alias set to obtain a first intersection; combining the drug medical advice recommendation set and the drug commodity name set to obtain an intersection to obtain a second intersection; the medical operation alias set and the medicine commodity name set are found based on a standard clinical path corresponding to a patient disease diagnosis name in the knowledge map and the patient characteristic information;
and the medical advice distribution module is used for acquiring the execution time of each medical advice in the first intersection and the second intersection through the second clinical path to obtain the medical advice distribution of each day/stage of the clinical path corresponding to the first intersection and the second intersection.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for automatic generation of a clinical pathway based on deep learning and knowledge-maps according to any one of claims 1 to 7.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps in the method for automatic generation of a deep learning and knowledgegraph-based clinical pathway of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010989064.4A CN112071431B (en) | 2020-09-18 | 2020-09-18 | Clinical path automatic generation method and system based on deep learning and knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010989064.4A CN112071431B (en) | 2020-09-18 | 2020-09-18 | Clinical path automatic generation method and system based on deep learning and knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112071431A true CN112071431A (en) | 2020-12-11 |
| CN112071431B CN112071431B (en) | 2024-02-23 |
Family
ID=73681228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010989064.4A Active CN112071431B (en) | 2020-09-18 | 2020-09-18 | Clinical path automatic generation method and system based on deep learning and knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112071431B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116994704A (en) * | 2023-09-22 | 2023-11-03 | 北斗云方(北京)健康科技有限公司 | Reasonable medication discrimination method based on clinical multi-modal data deep representation learning |
| CN117542535A (en) * | 2024-01-10 | 2024-02-09 | 电子科技大学 | Clinical path correction method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109378064A (en) * | 2018-10-29 | 2019-02-22 | 南京医基云医疗数据研究院有限公司 | Medical data processing method, device electronic equipment and computer-readable medium |
| CN109727680A (en) * | 2018-12-28 | 2019-05-07 | 上海列顿信息科技有限公司 | A kind of region clinical path management system based on big data technology |
| KR20200022110A (en) * | 2018-08-22 | 2020-03-03 | 주식회사 위담바이오 | Oriental medicine Clinical data collection and deep learning based data analysis system |
| CN111191020A (en) * | 2019-12-27 | 2020-05-22 | 江苏省人民医院（南京医科大学第一附属医院） | Prescription recommendation method and system based on machine learning and knowledge graph |
-
2020
- 2020-09-18 CN CN202010989064.4A patent/CN112071431B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20200022110A (en) * | 2018-08-22 | 2020-03-03 | 주식회사 위담바이오 | Oriental medicine Clinical data collection and deep learning based data analysis system |
| CN109378064A (en) * | 2018-10-29 | 2019-02-22 | 南京医基云医疗数据研究院有限公司 | Medical data processing method, device electronic equipment and computer-readable medium |
| CN109727680A (en) * | 2018-12-28 | 2019-05-07 | 上海列顿信息科技有限公司 | A kind of region clinical path management system based on big data technology |
| CN111191020A (en) * | 2019-12-27 | 2020-05-22 | 江苏省人民医院（南京医科大学第一附属医院） | Prescription recommendation method and system based on machine learning and knowledge graph |
Non-Patent Citations (1)
| Title |
|---|
| 崔洁;陈德华;乐嘉锦;: "基于EMR的乳腺肿瘤知识图谱构建研究", 计算机应用与软件, no. 12 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116994704A (en) * | 2023-09-22 | 2023-11-03 | 北斗云方(北京)健康科技有限公司 | Reasonable medication discrimination method based on clinical multi-modal data deep representation learning |
| CN116994704B (en) * | 2023-09-22 | 2023-12-15 | 北斗云方(北京)健康科技有限公司 | Reasonable medication discrimination method based on clinical multi-modal data deep representation learning |
| CN117542535A (en) * | 2024-01-10 | 2024-02-09 | 电子科技大学 | Clinical path correction method and device |
| CN117542535B (en) * | 2024-01-10 | 2024-03-22 | 电子科技大学 | Clinical path correction method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112071431B (en) | 2024-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111986770B (en) | Prescription medication auditing method, device, equipment and storage medium | |
| Ghosh et al. | Quro: facilitating user symptom check using a personalised chatbot-oriented dialogue system | |
| EP3234823B1 (en) | Differential medical diagnosis apparatus adapted in order to determine an optimal sequence of diagnostic tests for identifying a pathology by adopting diagnostic appropriateness criteria | |
| CN109920540A (en) | Construction method, device and the computer equipment of assisting in diagnosis and treatment decision system | |
| Kang et al. | Pretraining to recognize PICO elements from randomized controlled trial literature | |
| JP7304960B2 (en) | Health-informed prognostic score | |
| EP3928322A1 (en) | Automated generation of structured patient data record | |
| GB2586721A (en) | Adverse drug reaction analysis | |
| Raghavan et al. | emrkbqa: A clinical knowledge-base question answering dataset | |
| CN110689939A (en) | Recommendation method and device for medication sequence, readable medium and electronic equipment | |
| CN112071431B (en) | Clinical path automatic generation method and system based on deep learning and knowledge graph | |
| CN116805013A (en) | Traditional Chinese medicine video retrieval model based on knowledge graph | |
| CN112069329A (en) | Text corpus processing method, device, equipment and storage medium | |
| CN112259232B (en) | VTE risk automatic evaluation system based on deep learning | |
| Sarkar | Methods in biomedical informatics: a pragmatic approach | |
| CN117407682A (en) | Medical model evaluation method, device, electronic equipment and storage medium | |
| CN112966123A (en) | Medical health knowledge map system oriented to specific disease field | |
| Chen et al. | Automatically structuring on Chinese ultrasound report of cerebrovascular diseases via natural language processing | |
| CN111028951A (en) | Method and equipment for Chinese medicine diagnosis retrieval display and evaluation | |
| Nawroth | Supporting information retrieval of emerging knowledge and argumentation | |
| EP3937105A1 (en) | Methods and systems for user data processing | |
| Asif et al. | Using nanopublications to detect and explain contradictory research claims | |
| CN113643825A (en) | Medical case knowledge base construction method and system based on clinical key characteristic information | |
| EP4100957A1 (en) | Method and system for incorporating patient information | |
| Hussain et al. | Recommendation statements identification in clinical practice guidelines using heuristic patterns |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: No. 1237 Yingxiu Road, High tech Zone, Jinan City, Shandong Province, 250101 Applicant after: Zhongyang Health Technology Group Co.,Ltd. Address before: 12 / F, building 1, Aosheng building, 1166 Xinluo street, hi tech Zone, Jinan City, Shandong Province Applicant before: SHANDONG MSUNHEALTH TECHNOLOGY GROUP Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |