CN114822859A - Treatment thread mining and searching method and device - Google Patents
Treatment thread mining and searching method and device Download PDFInfo
- Publication number
- CN114822859A CN114822859A CN202210343899.1A CN202210343899A CN114822859A CN 114822859 A CN114822859 A CN 114822859A CN 202210343899 A CN202210343899 A CN 202210343899A CN 114822859 A CN114822859 A CN 114822859A
- Authority
- CN
- China
- Prior art keywords
- treatment thread
- treatment
- thread
- field
- target
- 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
- 238000011282 treatment Methods 0.000 title claims abstract description 431
- 238000000034 method Methods 0.000 title claims abstract description 93
- 238000005065 mining Methods 0.000 title claims abstract description 57
- 238000011272 standard treatment Methods 0.000 claims abstract description 126
- 239000003814 drug Substances 0.000 claims description 85
- 229940079593 drug Drugs 0.000 claims description 49
- 238000002560 therapeutic procedure Methods 0.000 claims description 30
- 238000012360 testing method Methods 0.000 claims description 21
- 206010067484 Adverse reaction Diseases 0.000 claims description 19
- 230000006838 adverse reaction Effects 0.000 claims description 19
- 238000012827 research and development Methods 0.000 claims description 17
- 238000003908 quality control method Methods 0.000 claims description 14
- 238000012544 monitoring process Methods 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 description 12
- 238000003745 diagnosis Methods 0.000 description 9
- 201000010099 disease Diseases 0.000 description 8
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 8
- 238000009093 first-line therapy Methods 0.000 description 7
- 206010050685 Cytokine storm Diseases 0.000 description 6
- 206010028980 Neoplasm Diseases 0.000 description 6
- 206010052015 cytokine release syndrome Diseases 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000009098 adjuvant therapy Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000003058 natural language processing Methods 0.000 description 5
- 238000002512 chemotherapy Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000011301 standard therapy Methods 0.000 description 4
- 206010041067 Small cell lung cancer Diseases 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 208000000587 small cell lung carcinoma Diseases 0.000 description 3
- 208000003950 B-cell lymphoma Diseases 0.000 description 2
- 206010059866 Drug resistance Diseases 0.000 description 2
- 238000011360 adjunctive therapy Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 201000011510 cancer Diseases 0.000 description 2
- 210000004027 cell Anatomy 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 238000010835 comparative analysis Methods 0.000 description 2
- 238000013479 data entry Methods 0.000 description 2
- 238000009509 drug development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012015 optical character recognition Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011369 optimal treatment Methods 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 238000009094 second-line therapy Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000011269 treatment regimen Methods 0.000 description 2
- MZOFCQQQCNRIBI-VMXHOPILSA-N (3s)-4-[[(2s)-1-[[(2s)-1-[[(1s)-1-carboxy-2-hydroxyethyl]amino]-4-methyl-1-oxopentan-2-yl]amino]-5-(diaminomethylideneamino)-1-oxopentan-2-yl]amino]-3-[[2-[[(2s)-2,6-diaminohexanoyl]amino]acetyl]amino]-4-oxobutanoic acid Chemical compound OC[C@@H](C(O)=O)NC(=O)[C@H](CC(C)C)NC(=O)[C@H](CCCN=C(N)N)NC(=O)[C@H](CC(O)=O)NC(=O)CNC(=O)[C@@H](N)CCCCN MZOFCQQQCNRIBI-VMXHOPILSA-N 0.000 description 1
- 208000017667 Chronic Disease Diseases 0.000 description 1
- 102100037850 Interferon gamma Human genes 0.000 description 1
- 108010074328 Interferon-gamma Proteins 0.000 description 1
- 108010002352 Interleukin-1 Proteins 0.000 description 1
- 108090001005 Interleukin-6 Proteins 0.000 description 1
- 208000032376 Lung infection Diseases 0.000 description 1
- 206010058467 Lung neoplasm malignant Diseases 0.000 description 1
- 108020001621 Natriuretic Peptide Proteins 0.000 description 1
- 102000004571 Natriuretic peptide Human genes 0.000 description 1
- 206010061309 Neoplasm progression Diseases 0.000 description 1
- 108060008682 Tumor Necrosis Factor Proteins 0.000 description 1
- 102000000852 Tumor Necrosis Factor-alpha Human genes 0.000 description 1
- 208000007502 anemia Diseases 0.000 description 1
- 230000001093 anti-cancer Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000009108 consolidation therapy Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000011157 data evaluation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000011354 first-line chemotherapy Methods 0.000 description 1
- 238000001794 hormone therapy Methods 0.000 description 1
- 238000011221 initial treatment Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 201000005202 lung cancer Diseases 0.000 description 1
- 208000020816 lung neoplasm Diseases 0.000 description 1
- 238000009115 maintenance therapy Methods 0.000 description 1
- 230000002503 metabolic effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000000692 natriuretic peptide Substances 0.000 description 1
- 238000009099 neoadjuvant therapy Methods 0.000 description 1
- 238000009116 palliative therapy Methods 0.000 description 1
- 230000007170 pathology Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000001959 radiotherapy Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000009118 salvage therapy Methods 0.000 description 1
- 238000011519 second-line treatment Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 238000009095 third-line therapy Methods 0.000 description 1
- UCFGDBYHRUNTLO-QHCPKHFHSA-N topotecan Chemical compound C1=C(O)C(CN(C)C)=C2C=C(CN3C4=CC5=C(C3=O)COC(=O)[C@]5(O)CC)C4=NC2=C1 UCFGDBYHRUNTLO-QHCPKHFHSA-N 0.000 description 1
- 229960000303 topotecan Drugs 0.000 description 1
- 230000005751 tumor progression Effects 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- 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
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
-
- 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
- G16H70/00—ICT specially adapted for the handling or processing of medical references
- G16H70/40—ICT specially adapted for the handling or processing of medical references relating to drugs, e.g. their side effects or intended usage
Abstract
The invention provides a method and a device for mining and searching a treatment thread, wherein the method for mining the treatment thread comprises the following steps: determining text associated with a treatment thread; performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field; and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text. The treatment thread mining and retrieving method and device provided by the embodiment of the invention can realize comprehensive and reliable treatment thread mining, effectively improve the realization efficiency of the treatment thread mining and reduce the cost of the treatment thread mining.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method and a device for mining and searching a treatment thread.
Background
The treatment thread is a description of the treatment sequence in the disease treatment process. Determining treatment threads relates to optimal treatment option selection, drug resistance management, and confidence in benefit-risk. In the precise treatment era, the clinical positioning of the diagnosis and treatment of patients is more and more careful, and the management requirements on treatment threads are higher and higher.
The current treatment process is widely applied clinically, but is not reflected in medical knowledge or products. The determination of the treatment thread needs to manually review various medical texts, analyze and judge, has low efficiency and certain subjective judgment, and is easy to miss key information.
Disclosure of Invention
The invention provides a method and a device for mining and searching a treatment thread, which are used for solving the defects that the efficiency is low and key information is easy to miss when analysis and judgment are carried out manually in the prior art.
The invention provides a treatment thread mining method, which comprises the following steps:
determining text associated with a treatment thread;
performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field;
and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
According to the treatment thread mining method provided by the invention, the standard treatment thread matched with the treatment thread field is determined based on the semantics of the treatment thread field and the semantics of the standard treatment thread field in a treatment thread dictionary, and the method comprises the following steps:
carrying out semantic coding on the treatment thread field to obtain semantic features of the treatment thread field, and calculating semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field;
and/or inputting the treatment thread field and the standard treatment thread field into a semantic sorting model to obtain a matching score between the treatment thread field and the standard treatment thread field output by the semantic sorting model;
determining a standard treatment thread matching the treatment thread field based on the semantic similarity and/or the match score.
According to the treatment thread mining method provided by the invention, the step of inputting the treatment thread field and the standard treatment thread field into a semantic sorting model to obtain the matching score between the treatment thread field and the standard treatment thread field output by the semantic sorting model comprises the following steps:
determining a plurality of standard treatment threads related to the treatment thread field based on the semantic similarity;
respectively inputting the plurality of standard treatment thread fields and the treatment thread fields into the semantic sorting model to obtain a plurality of matching scores output by the semantic sorting model;
the determining a standard treatment thread matching the treatment thread field based on the semantic similarity and/or the match score comprises:
and sequencing the plurality of matching scores from high to low, and determining the standard treatment thread corresponding to the matching score which is sequenced first and is greater than or equal to a preset threshold as the standard treatment thread matched with the treatment thread field.
According to the treatment thread mining method provided by the invention, the determination of the text related to the treatment thread comprises the following steps:
determining a target medical text comprising at least one of a clinical trial result class text, a treatment class guideline text, and a treatment class clinical trial text;
and carrying out treatment thread relevance classification on the target medical text to obtain a text related to a treatment thread.
The invention provides a retrieval method, which comprises the following steps:
receiving a target retrieval keyword sent by a user terminal, wherein the target retrieval keyword comprises a target treatment thread and/or a target medicine;
determining medical data associated with the target retrieval keyword from predetermined medical data of each medicine under each treatment thread;
each treatment thread corresponding to the medical data is determined based on the treatment thread mining method as described in any one of the above, and the medical data includes at least one of guideline data, clinical trial data and epidemiological data.
According to the search method provided by the present invention, in a case where the target search keyword includes the target drug, the determining, from the predetermined medical data of each drug in each treatment thread, the medical data associated with the target search keyword includes:
and determining clinical test data of the indications corresponding to the target medicines under each treatment thread from predetermined medical data of the medicines under each treatment thread, wherein the clinical test data comprises at least one of a medicine research and development stage, a highest research and development stage of the indications corresponding to the target medicines and an approval mechanism for marketing the medicines.
According to the search method provided by the present invention, in a case where the target search keyword includes the target treatment thread, the method further includes the step of determining medical data associated with the target search keyword from predetermined medical data of each drug under each treatment thread, and then:
acquiring medical record data of a patient;
structuring the patient medical record data to obtain structured medical record data;
and monitoring the patient medical record data based on the structured medical record data and the medical record medical quality control index threshold value under the target treatment thread.
According to the retrieval method provided by the invention, the monitoring of the patient medical record data based on the structured medical record data and the medical record medical quality control index threshold under the target treatment thread comprises the following steps:
and estimating adverse reaction data corresponding to the patient medical record data under the target treatment thread based on the structured data of the risk factors contained in the structured medical record data and the predetermined relationship between the adverse reaction of different drug treatments and the risk factors generating the adverse reaction.
The invention also provides a therapeutic thread mining device, comprising:
the text determining unit is used for determining texts related to the treatment threads;
the entity identification unit is used for carrying out treatment thread entity identification on the text and taking an entity obtained by identification as a treatment thread field;
and the treatment thread determining unit is used for determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
The present invention also provides a retrieval apparatus comprising:
the system comprises a keyword receiving unit, a target searching unit and a target searching unit, wherein the keyword receiving unit is used for receiving a target searching keyword sent by a user terminal, and the target searching keyword comprises a target treatment thread and/or a target medicine;
the medical data determining unit is used for determining medical data associated with the target retrieval keyword from the predetermined medical data of each medicine under each treatment thread; each treatment thread in the medical data is determined based on a treatment thread mining method as described in any of the above, the medical data comprising at least one of guideline data, clinical trial data, and epidemiological data.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the therapy thread mining method or the retrieval method as described in any one of the above.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a therapy thread mining method or a retrieval method as described in any of the above.
The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a method of therapy thread mining or retrieval as described in any one of the above.
The treatment thread mining and searching method and the device provided by the invention are used for carrying out treatment thread entity identification on texts related to treatment threads, and using the entity obtained by identification as a treatment thread field; and determining the standard treatment thread matched with the treatment thread field according to the semantics of the treatment thread field and the semantics of the standard treatment thread field in the treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text. The method effectively improves the realization efficiency of the treatment thread mining and reduces the cost of the treatment thread mining while realizing comprehensive and reliable treatment thread mining.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a therapy thread mining method provided by the present invention;
FIG. 2 is a flow chart of a therapy thread matching method according to the present invention;
FIG. 3 is a second flowchart of the method for matching treatment threads according to the present invention;
FIG. 4 is a flow chart illustrating a text acquisition method associated with a treatment thread according to the present invention;
FIG. 5 is a schematic flow chart of a retrieval method provided by the present invention;
FIG. 6 is a flow chart of a method for monitoring patient medical record data according to the present invention;
FIG. 7 is a schematic structural diagram of a treatment thread mining device according to the present invention;
FIG. 8 is a schematic structural diagram of a search device provided in the present invention;
fig. 9 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. 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 invention.
The treatment of various chronic diseases, especially tumor diseases, is often long and complicated, and it is necessary to administer targeted treatment in stages, such as a treatment scheme of first-stage treatment, a treatment scheme of second-stage treatment, and adjuvant treatment, last-line treatment, etc., according to the pathology, stage, physical condition, and existing medication of a patient.
Determining treatment threads relates to optimal treatment option selection, drug resistance management, and confidence in benefit-risk. In the precise treatment era, the clinical positioning of the diagnosis and treatment of patients is more and more careful, and the management requirements on treatment threads are higher and higher.
However, the current treatment process has a disjointed phenomenon in clinical application and medical texts, so that the treatment process is widely applied in clinical application on one hand, but is not embodied in medical knowledge or products on the other hand, and even the concept of the treatment process is not mentioned in the medical college textbooks.
The pharmaceutical field also has obvious disjunction with clinical application, for example, when a medicine is registered in a clinical test, various conditions are defined in the indication or inclusion and exclusion standard of the medicine, but the treatment process is not mentioned, which is a ubiquitous realistic problem in pharmaceutical knowledge and databases. When reading the pharmaceutical information, the industry user needs to manually analyze and judge to correspond to the clinical treatment scheme.
The manual review mode has low efficiency and certain subjective judgment, key information is easy to miss, and how to realize the quick searching and positioning of the treatment thread becomes the problem to be solved at present. In view of the above problems, embodiments of the present invention provide a method for mining a treatment thread, which is used for mining a treatment thread in a medical text to achieve quick search and location of the treatment thread.
Fig. 1 is a schematic flow chart of a therapy thread mining method provided by the present invention, and as shown in fig. 1, the method includes:
at step 110, text associated with the treatment thread is determined.
Here, the text related to the treatment thread is a partial paragraph or a whole paragraph in the medical text that needs to be mined by the treatment thread, and the text related to the treatment thread may be obtained by crawling from a related website by using a web crawler, or obtained by performing image shooting or scanning on a paper version of the medical text, and the embodiment of the present invention is not particularly limited to this.
The acquisition of text associated with a treatment thread may be a timed acquisition or a real-time monitoring of the corresponding text-posting website, such as a real-time monitoring of updates or revision changes to the clinical guideline, and acquiring clinical guideline data for a new version of the clinical guideline after the new version of the clinical guideline appears. Sources of text related to a treatment thread include, but are not limited to: journal literature, clinical guideline data, patient medical history, clinical trial registration information, and the like.
And step 120, performing treatment thread entity identification on the text, and taking the entity obtained by identification as a treatment thread field.
Specifically, the treatment thread field herein is a field capable of representing a treatment thread, such as first-line treatment, second-level treatment, or adjuvant treatment. For the text related to the treatment thread acquired in step 110, in which only a part of the fields are fields representing the treatment thread, in order to more accurately obtain the fields representing the treatment thread, it is necessary to perform treatment thread entity identification on the text, and the identified entities are used as the treatment thread fields.
To implement therapy thread entity recognition for text, an entity recognition model may be trained in advance. The entity referred to here, including the treatment thread field for the text related to the treatment thread, may input the text into a pre-trained entity recognition model, perform treatment thread entity recognition on the text by the entity recognition model, and output an entity label of each word in the entity text, where the label system of the entity recognition may be BIO, biees, etc., B represents the beginning of the entity, E represents the end of the entity, I represents the middle word of the entity, O represents a non-entity, and S represents a single entity.
Prior to execution of step 120, the entity recognition model may be trained. The training method of the entity recognition model can comprise the following steps: firstly, a large amount of sample texts related to treatment threads are collected, and entities and entity types in the sample texts are labeled manually. And then, training the initial entity recognition model based on the sample text and the entity and entity type marked in the sample text, thereby obtaining the entity recognition model.
And step 130, determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of the standard treatment thread field in the treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as a treatment thread corresponding to the text.
Specifically, the entity obtained by identification is used as a treatment thread field, and on this basis, a standard treatment thread field matched with the treatment thread field can be selected according to the semantics of the treatment thread field and used as a treatment thread corresponding to the text. The standard treatment thread field matched with the treatment thread field can be understood as that the treatment thread field obtained by entity recognition is greatly overlapped with the standard treatment thread field, or the semantics of the treatment thread field is greatly overlapped with the semantics of the standard treatment thread field.
Here, the therapy thread dictionary includes standard therapy thread fields that are used more clinically, and the standard therapy thread fields may be different for different fields of disease. The construction of the treatment thread dictionary can be realized by searching a large amount of medical literature, summarizing and analyzing, determining the classification standard and definition of the treatment threads in the main disease field, and specifically can be realized by Natural Language Processing (NLP) or manual work, or a combination of NLP and manual work.
Further, the standard therapy thread fields in the therapy thread dictionary include, but are not limited to: neoadjuvant therapy, adjuvant therapy, initial therapy, induction therapy, consolidation therapy, maintenance therapy, first line therapy, second line therapy, third line therapy, last line therapy, rescue therapy, palliative therapy, and other types of therapy.
By way of example in the tumour area, first line therapy refers to the first treatment cycle after diagnosis, when the treatment regime is best and has minimal side effects, also referred to as primary treatment or therapy. The goal of first-line therapy is to cure cancer where possible. Such as: the EP regimen for treating small cell lung cancer is a classic first-line treatment regimen, and this group of treatments allows the control of most small cell lung cancers and even the complete disappearance of tumor lesions.
Second line therapy refers to patients who have re-developed tumor progression after first line therapy and are resistant to first line therapy regimens requiring replacement of regimens with different anti-cancer mechanisms. Compared with the first-line treatment, the second-line treatment scheme has the disadvantages of first-line treatment, larger side effect or higher price. Small cell lung cancer is also exemplified: after failure of the first-line EP regimen, the second-line recommended treatment regimen is topotecan, which is also the classical second-line recommendation.
Adjuvant therapy is also referred to as adjunctive therapy. Usually post-operatively, to destroy any remaining cancer cells in the body, and adjunctive therapy to reduce the likelihood of tumor recurrence or dissemination elsewhere. The adjuvant therapy may include radiation therapy, chemotherapy, hormonal therapy, or further surgical therapy.
Considering that the treatment thread field obtained by entity recognition may or may not be a standard treatment thread field, a standard treatment thread matching the treatment thread field may be further determined according to the matching degree of the semantics of the treatment thread field and the semantics of the standard treatment thread field, and the standard treatment thread matching the treatment thread field is determined as the treatment thread corresponding to the text.
According to the treatment thread mining method provided by the embodiment of the invention, treatment thread entity identification is carried out on texts related to treatment threads, and the entity obtained by identification is used as a treatment thread field; and determining the standard treatment thread matched with the treatment thread field according to the semantics of the treatment thread field and the semantics of the standard treatment thread field in the treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text. The method effectively improves the realization efficiency of the treatment thread mining and reduces the cost of the treatment thread mining while realizing comprehensive and reliable treatment thread mining.
Based on any of the above embodiments, fig. 2 is a schematic flowchart of a method for matching a treatment thread provided by the present invention, and as shown in fig. 2, in step 130, a standard treatment thread matching a treatment thread field is determined based on semantics of the treatment thread field and semantics of the standard treatment thread field in a treatment thread dictionary, including:
131, carrying out semantic coding on the treatment thread field to obtain the semantic features of the treatment thread field, and calculating the semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field;
based on the semantic similarity and/or the match score, a standard treatment thread matching the treatment thread field is determined, step 133.
Specifically, semantically encoding the therapy thread field may map the therapy thread field to a vector space, convert it into a digitally represented word-embedded representation vector, and thereby obtain a semantic representation vector, i.e., a semantic feature, capable of characterizing the therapy thread field. Semantic features may be output as a result in a vector representation.
And after the semantic features of the treatment thread field are obtained, calculating the semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field. Specifically, the similarity may be implemented by a similarity algorithm, such as cosine similarity, euclidean distance, or minkowski distance, which is not specifically limited in this embodiment of the present invention.
The larger the calculated semantic similarity value is, the higher the matching degree of the treatment thread field and the standard treatment line segment field is; the smaller the semantic similarity value obtained by calculation, the lower the matching degree of the treatment thread field and the standard treatment line segment field is.
In step 132, the semantic sorting model is used to calculate a matching degree between the treatment thread field and the standard treatment thread field, i.e. a degree to which the treatment thread field can represent the standard treatment thread. And inputting the treatment thread field and the standard treatment thread field into the semantic sorting model to obtain a matching score between the treatment thread field and the standard treatment thread field output by the semantic sorting model. The higher the output matching score is, the higher the matching degree of the treatment thread field and the standard treatment line segment field is; the lower the match score output, the lower the degree of match of the therapy thread field with the standard therapy segment field.
Then, the corresponding standard treatment thread with the highest semantic similarity can be determined as the standard treatment thread matched with the treatment thread field only according to the calculated semantic similarity; the matching score output by the semantic sorting model can be only used, for example, the standard treatment thread corresponding to the highest matching score is determined as the standard treatment thread matched with the treatment thread field; the semantic similarity and the matching score can also be considered comprehensively, for example, a plurality of candidate standard treatment threads are determined according to the semantic similarity, and then the standard treatment thread matched with the treatment thread field is determined from the plurality of candidate standard treatment threads according to the matching score. The standard treatment thread determining method matched with the treatment thread field can be flexibly selected according to actual conditions, and the embodiment of the invention is not particularly limited in this respect.
According to the treatment thread mining method provided by the embodiment of the invention, the standard treatment thread matched with the treatment thread field is determined based on the semantic similarity and/or the matching score, so that comprehensive and reliable treatment thread mining can be further realized.
Based on any of the above embodiments, fig. 3 is a second schematic flow chart of the therapy thread matching method provided by the present invention, as shown in fig. 3, step 132 specifically includes:
step 132-1, determining a plurality of standard treatment threads related to the treatment thread field based on the semantic similarity;
step 132-2, inputting the plurality of standard treatment thread fields and the treatment thread fields into the semantic sorting model respectively to obtain a plurality of matching scores output by the semantic sorting model;
correspondingly, step 133 specifically includes:
and step 133-1, sorting the plurality of matching scores in a descending order, and determining the standard treatment thread corresponding to the matching score which is sorted first and is greater than or equal to a preset threshold as the standard treatment thread matched with the treatment thread field.
Specifically, after calculating the semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field, on the basis, according to the semantic similarity, determining a plurality of standard treatment threads related to the treatment thread field. The plurality of standard treatment threads can be a plurality of standard treatment threads with semantic similarity with the treatment thread field being greater than or equal to a preset threshold; or a preset threshold number of standard treatment threads with the semantic similarity ranking at the top, such as 10 standard treatment threads.
And obtaining a plurality of standard treatment threads through semantic similarity screening, and then respectively inputting a plurality of standard treatment thread fields and treatment thread fields into the semantic sorting model to obtain a plurality of matching scores output by the semantic sorting model.
And sequencing the plurality of standard treatment threads according to the sequence of the matching scores from high to low, determining the standard treatment thread which is sequenced first and has the matching score larger than or equal to a preset threshold as the standard treatment thread matched with the treatment thread field, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
It should be noted that if the first-ranked match score does not reach the preset threshold, it indicates that the text may not have a corresponding treatment thread, but the treatment thread corresponding to the first-ranked match score may still be used as the most likely treatment thread of the text.
Based on any of the above embodiments, fig. 4 is a schematic flowchart of a text acquisition method related to a treatment thread provided by the present invention, as shown in fig. 4, step 110 specifically includes:
and 112, performing treatment thread relevance classification on the target medical text to obtain a text related to the treatment thread.
Specifically, in consideration of more sources of medical texts, in order to further improve the efficiency of therapy thread mining, the medical texts may be classified by a text classifier, and a target medical text is obtained by screening, where the target medical text includes at least one of a clinical test result type text, a therapy type guideline text, and a therapy type clinical test text.
The clinical test result type text means that the clinical test text not only discloses candidate drugs, test endpoints and specific embodiments of clinical tests, but also discloses obtained clinical test data which are related to the clinical test endpoints and can be used for evaluating the metabolic characteristics, safety and effectiveness of the drugs in human bodies.
The treatment class guideline text is a text describing guidelines related to clinical treatment protocols, such as clinical practice guidelines, standard of care (SOC), and related documents.
The treatment-type clinical trial text is a text in which a text relating to a treatment-type clinical trial, for example, a text relating to interventional clinical trial information registered by a clinical trial registration platform of each country is described.
The source format of the target medical text can be a plain text format or a rich text format. Aiming at the plain text format, the plain text format can be obtained by directly reading the text; for the rich text format, the text may be further processed, such as performing text Recognition on the picture by Optical Character Recognition (OCR), or performing text extraction by pdftotext.
Considering that the target medical texts have different data distributions, different sources, usually long and complicated texts, and many texts that do not contain the content of the "treatment thread", the target medical texts need to be further classified according to whether the classification is related to the treatment thread or not. Furthermore, treatment thread relevance classification can be carried out on each paragraph in the target medical text, and the paragraphs relevant to the treatment thread are obtained.
In order to implement the treatment thread relevance classification for the text, the text can be distinguished by adopting a pre-trained binary classification model, so that the text containing the treatment thread, namely the text related to the treatment thread, is identified.
According to the method provided by the embodiment of the invention, the text related to the treatment thread can be conveniently and quickly acquired by carrying out treatment thread relevance classification on the target medical text.
Based on any of the above embodiments, fig. 5 is a schematic flow chart of the retrieval method provided by the present invention, and as shown in fig. 5, the method includes:
Specifically, the medical texts from various sources can be mined through the treatment threads by the method described in the above embodiment, so as to obtain medical data of each medicine under each treatment thread, thereby providing convenience for information search and positioning. And on the basis, a retrieval system can be constructed so as to facilitate the quick search and positioning of the target information.
The user can input the target search keyword through a user terminal in the form of a smart phone, a computer, a tablet computer and the like, and the target search keyword is sent to a server for searching. The target search keywords herein are keywords that are expected to find relevant information from medical data, and may include, for example, a target treatment thread and/or a target drug.
each treatment thread corresponding to the medical data is determined based on the treatment thread mining method, and the medical data comprises at least one of guideline data, clinical trial data and epidemiological data.
Specifically, after receiving the target search keyword, the medical data associated with the target search keyword can be searched and located in the predetermined medical data of each drug under each treatment thread, where the medical data includes at least one of guideline data, clinical trial data, and epidemiological data.
The guideline data may be information such as an indication name, a recommended treatment medicine corresponding to the indication, a clinical stage, a recommended level, a special medication crowd and the like acquired based on the guideline.
The clinical trial data may be information such as a trial group drug name, a control group drug name, an indication name, a trial result, and the like acquired based on the clinical trial registration platforms of each country and clinical trial literature.
The epidemiological data can be information such as the name of the indication, the number of new cases, the number of dead cases, the total number of patients with diseases and the like in the corresponding calendar year of the indication, which are acquired based on related epidemiological documents.
The indication name and the medicine name may be standardized indication names and medicine names, and the specific processing mode may be obtained by matching in a pre-constructed medicine dictionary or an indication dictionary.
According to the retrieval method provided by the embodiment of the invention, the medical data associated with the target medicine and/or the target treatment thread is quickly retrieved through the medical data of each medicine under each treatment thread, the information query efficiency is improved, and the unified comparative analysis based on the treatment thread dimension can be realized on each medical data.
Based on any of the above embodiments, in the case that the target search keyword includes the target drug, step 520 specifically includes:
and determining clinical test data of the indications corresponding to the target medicines under each treatment thread from the predetermined medical data of the medicines under each treatment thread, wherein the clinical test data comprises at least one of a medicine research and development stage, a highest research and development stage of the indications corresponding to the target medicines and an approval mechanism for marketing the medicines.
Specifically, in the case that the target search keyword includes a target drug, a treatment thread information set and an indication information set corresponding to the target drug may be obtained from predetermined medical data of each drug under each treatment thread according to the drug, the indication, the association relationship between the treatment thread and the development stage.
Further, acquiring a research and development stage information set corresponding to all corresponding medicine indications under the target treatment thread; acquiring a research and development stage information set of treatment threads corresponding to all medicines under a target indication; and acquiring a target treatment thread and a corresponding drug development stage information set under the target indication.
In one embodiment, the drug development phase information set may be represented as a base table of m x n, with each element in the table being a development phase.
The weight is given to each research and development stage one by one, along with the progress of the research and development stages, the weight value corresponding to each research and development stage is gradually increased, the probability that the conversion and upgrade of each research and development stage are successful or failed is preset to be 50%, and therefore the weight difference of two adjacent research and development stages is set to be 0.5. A secondary table of m x n is generated.
For example: the stages of development include "preclinical", "declared clinical", "phase I/II clinical", "phase II/III clinical", "phase III/IV clinical", "phase IV clinical", "application to market" and "approval to market", each stage of development corresponding to a weight of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0 and 5.5, respectively.
And performing row sorting and column sorting on the generated secondary table, wherein the sorting mode is as follows: the descending order of the sum of the weights of each row or each column generates the final m x n table.
Furthermore, the highest research and development stage of the indication corresponding to the target drug can be obtained, and if the highest research and development stage is listed, an approval mechanism for listing the drug is further obtained and provided for the user to perform further screening.
Based on any of the above embodiments, fig. 6 is a schematic flowchart of a patient medical record data monitoring method provided by the present invention, and as shown in fig. 6, in a case that the target search keyword includes a target treatment thread, the step 520 further includes:
Specifically, on the basis of the retrieval system, medical quality control, risk assessment, clinical scientific research and data insight based on real medical record data of a patient can be realized according to the dimension of a treatment thread.
The medical record data of the patient is mainly in a text form, a picture form or a video form, and is not limited specifically herein.
The patient medical record Data is then Structured, the Structured methods including Natural Language Processing (NLP) and Structured Data Entry (Structured Data Entry). Structured medical record data can be obtained, including but not limited to, patient symptoms, signs, medical history, diagnosis, treatment threads, exam examinations, medication orders, and the like.
According to the diagnosis and treatment guidelines of diseases or national diagnosis and treatment specifications, a standard disease diagnosis and treatment path is split. And determining medical record medical quality control indexes according to the disease diagnosis and treatment path. For example, first-line chemotherapy indicators for patients with advanced lung cancer include: the electrocardiogram completion rate after 24 hours of hospital admission, the inspection completion rate after 24 hours of hospital admission, whether anemia exists before chemotherapy, whether leucocyte is reduced before chemotherapy, the first-line treatment drug dose adjustment rate, the lung infection rate after chemotherapy and the like. And then determining a calculation method of the medical quality control index, and automatically calculating by using the structured medical record data to monitor the medical record data of the patient.
And the early warning threshold value can be determined according to historical data and clinical quality control experience, and when the data of related medical quality control indexes in the medical record data reaches or exceeds the threshold value, automatic early warning is performed.
The retrieval method provided by the embodiment of the invention realizes the quality control of medical indexes of the medical record data of the patient from the dimension of the treatment thread, and improves the medical record management efficiency.
Based on any of the above embodiments, step 630 specifically includes:
and estimating adverse reaction data corresponding to the patient medical record data under the target treatment thread based on the structured data of each risk factor contained in the structured medical record data and the predetermined relationship between the adverse reaction of different drug treatments and each risk factor generating the adverse reaction.
Specifically, on the basis of medical record quality control, adverse reaction data corresponding to patient medical record data in a target thread can be estimated.
Firstly, main complications or adverse reactions of different drug treatments and potential risk factors of the complications or adverse reactions are determined according to literature reports and diagnosis and treatment guidelines of diseases. For example, cytokine storm is a common adverse reaction of last-stage B cell lymphoma treated by Cart-T cell end line, serious cytokine storm is very dangerous, and possible risk factors comprise age, sex, treatment process, past medication history, IL-1/IL-6/TNF-alpha/IFN-gamma concentration in blood, natriuretic peptide concentration and the like.
Then extracting the structured data of each risk factor in the medical record data of the patient with the severe cytokine storm, and calculating the relationship between the severe cytokine storm and different risk factors by a statistical method of multivariate correlation analysis.
Therefore, according to the structured data of each risk factor contained in the medical record data of the B cell lymphoma patient and the relationship between the severe cytokine storm and different risk factors, the data of adverse reaction of the severe cytokine storm occurring in the medical record data of the patient treated by using Car-T can be evaluated.
According to the method provided by the embodiment of the invention, the risk assessment of the adverse reaction corresponding to the patient medical record data is realized from the dimension of the treatment thread, so that the refined treatment is realized.
Based on any embodiment, a retrieval method based on a therapy thread dimension is provided, which includes:
s1, acquiring a target medical text, wherein the target medical text comprises at least one of a clinical test result type text, a treatment type guideline text and a treatment type clinical test text.
And S2, performing treatment thread relevance classification on the target medical text by adopting a binary classification model to obtain a text related to a treatment thread.
S3, performing treatment thread entity recognition on the text related to the treatment thread, and taking the recognized entity as a treatment thread field;
s4, carrying out semantic coding on the treatment thread field to obtain the semantic features of the treatment thread field, and calculating the semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field;
s5, determining a plurality of standard treatment threads related to the treatment thread field based on the semantic similarity; and respectively inputting the plurality of standard treatment thread fields and the treatment thread fields into the semantic sorting model to obtain a plurality of matching scores output by the semantic sorting model.
S6, the matching scores are sorted from high to low, and the standard treatment thread corresponding to the matching score which is sorted first and is greater than or equal to a preset threshold is determined as the standard treatment thread matched with the treatment thread field. And determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
And S7, acquiring a treatment thread corresponding to each medical text according to the steps S1-S6, and acquiring medical data of each medicine under each treatment thread, wherein the medical data comprises at least one of guideline data, clinical trial data and epidemiological data.
S8, receiving a target treatment thread and/or a target medicine sent by a user terminal; determining medical data associated with the target treatment thread and/or target medicine from the medical data of each medicine determined in S7 under each treatment thread; and returning the medical data to the user terminal.
In addition, the medical index quality control of the patient medical record data and the adverse reaction data evaluation corresponding to the patient medical record data can be realized according to the dimension of the treatment thread.
The following describes the treatment thread mining apparatus provided by the present invention, and the treatment thread mining apparatus described below and the treatment thread mining method described above may be referred to correspondingly.
Based on any of the above embodiments, fig. 7 is a schematic structural diagram of a treatment thread mining device provided by the present invention, as shown in fig. 7, the device includes:
a text determination unit 710 for determining text related to the treatment thread;
an entity identification unit 720, configured to perform therapy thread entity identification on the text, and use an entity obtained through identification as a therapy thread field;
a treatment thread determining unit 730, configured to determine, based on the semantics of the treatment thread field and the semantics of the standard treatment thread field in the treatment thread dictionary, a standard treatment thread matching the treatment thread field, and determine the standard treatment thread matching the treatment thread field as the treatment thread corresponding to the text.
The treatment thread mining device provided by the embodiment of the invention identifies the entity of the treatment thread for the text related to the treatment thread, and takes the entity obtained by identification as the field of the treatment thread; and determining the standard treatment thread matched with the treatment thread field according to the semantics of the treatment thread field and the semantics of the standard treatment thread field in the treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text. The method effectively improves the realization efficiency of the treatment thread mining and reduces the cost of the treatment thread mining while realizing comprehensive and reliable treatment thread mining.
Based on any of the above embodiments, the treatment thread determining unit 730 is further configured to:
carrying out semantic coding on the treatment thread field to obtain semantic features of the treatment thread field, and calculating semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field;
and/or inputting the treatment thread field and the standard treatment thread field into a semantic sorting model to obtain a matching score between the treatment thread field and the standard treatment thread field output by the semantic sorting model;
determining a standard treatment thread matching the treatment thread field based on the semantic similarity and/or the match score.
Based on any of the above embodiments, the treatment thread determining unit 730 is further configured to:
determining a plurality of standard treatment threads related to the treatment thread field based on the semantic similarity;
respectively inputting the plurality of standard treatment thread fields and the treatment thread fields into the semantic sorting model to obtain a plurality of matching scores output by the semantic sorting model;
and sequencing the multiple matching scores from high to low, and determining the standard treatment thread corresponding to the matching score which is the first in sequencing and is greater than or equal to a preset threshold as the standard treatment thread matched with the treatment thread field.
Based on any of the above embodiments, the text determination unit 710 is further configured to:
determining a target medical text comprising at least one of a clinical trial result class text, a treatment class guideline text, and a treatment class clinical trial text;
and carrying out treatment thread relevance classification on the target medical text to obtain a text related to a treatment thread.
Based on any of the above embodiments, fig. 8 is a schematic structural diagram of a retrieval apparatus provided by the present invention, and as shown in fig. 8, the retrieval apparatus includes:
a keyword receiving unit 810, configured to receive a target search keyword sent by a user terminal, where the target search keyword includes a target treatment thread and/or a target drug;
a medical data determining unit 820, configured to determine, from predetermined medical data of each drug under each treatment thread, medical data associated with the target search keyword; each treatment thread corresponding to the medical data is determined based on the treatment thread mining method as described in any one of the above, and the medical data includes at least one of guideline data, clinical trial data and epidemiological data.
The retrieval device provided by the embodiment of the invention realizes the rapid retrieval of the medical data associated with the target medicine and/or the target treatment thread through the medical data of each medicine under each treatment thread, is beneficial to improving the information query efficiency, and can realize the unified comparative analysis of each medical data based on the treatment thread dimension.
Based on any of the above embodiments, in a case that the target search keyword includes the target drug, the medical data determining unit 820 is further configured to:
and determining clinical test data of the indications corresponding to the target medicines under each treatment thread from predetermined medical data of the medicines under each treatment thread, wherein the clinical test data comprises at least one of a medicine research and development stage, a highest research and development stage of the indications corresponding to the target medicines and an approval mechanism for marketing the medicines.
Based on any embodiment of the foregoing, in a case that the target search keyword includes the target treatment thread, the search apparatus further includes a data monitoring unit, configured to:
acquiring medical record data of a patient;
structuring the patient medical record data to obtain structured medical record data;
and monitoring the patient medical record data based on the structured medical record data and the medical record medical quality control index threshold value under the target treatment thread.
Based on any of the above embodiments, the data monitoring unit is further configured to:
and estimating adverse reaction data corresponding to the patient medical record data under the target treatment thread based on the structured data of the risk factors contained in the structured medical record data and the predetermined relationship between the adverse reaction of different drug treatments and the risk factors generating the adverse reaction.
Fig. 9 illustrates a physical structure diagram of an electronic device, and as shown in fig. 9, the electronic device may include: a processor (processor)910, a communication Interface (Communications Interface)920, a memory (memory)930, and a communication bus 940, wherein the processor 910, the communication Interface 920, and the memory 930 communicate with each other via the communication bus 940. Processor 910 may invoke logic instructions in memory 930 to perform a therapy thread mining method comprising: determining text associated with a treatment thread; performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field; and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
Furthermore, the logic instructions in the memory 930 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being storable on a non-transitory computer-readable storage medium, the computer program, when executed by a processor, being capable of executing the method for therapy thread mining provided by the above methods, the method comprising: determining text associated with a treatment thread; performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field; and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
When the computer program is executed by a processor, the computer can execute the searching method provided by the methods, and the method comprises the following steps: receiving a target retrieval keyword sent by a user terminal, wherein the target retrieval keyword comprises a target treatment thread and/or a target medicine; determining medical data associated with the target retrieval keyword from predetermined medical data of each medicine under each treatment thread; each treatment thread corresponding to the medical data is determined based on the treatment thread mining method, and the medical data comprises at least one of guideline data, clinical trial data and epidemiological data.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of therapy thread mining provided by the above methods, the method comprising: determining text associated with a treatment thread; performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field; and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
When the computer program is executed by a processor, the computer can execute the searching method provided by the methods, and the method comprises the following steps: receiving a target retrieval keyword sent by a user terminal, wherein the target retrieval keyword comprises a target treatment thread and/or a target medicine; determining medical data associated with the target retrieval keyword from predetermined medical data of each medicine under each treatment thread; each treatment thread corresponding to the medical data is determined based on the treatment thread mining method, and the medical data comprises at least one of guideline data, clinical trial data and epidemiological data.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on the understanding, the above technical solutions substantially or otherwise contributing to the prior art may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the various embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method of therapy thread mining, comprising:
determining text associated with a treatment thread;
performing treatment thread entity identification on the text, and taking an entity obtained by identification as a treatment thread field;
and determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
2. The method of claim 1, wherein determining a standard treatment thread that matches the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary comprises:
carrying out semantic coding on the treatment thread field to obtain semantic features of the treatment thread field, and calculating semantic similarity between the semantic features of the treatment thread field and the semantic features of the standard treatment thread field;
and/or inputting the treatment thread field and the standard treatment thread field into a semantic sorting model to obtain a matching score between the treatment thread field and the standard treatment thread field output by the semantic sorting model;
determining a standard treatment thread matching the treatment thread field based on the semantic similarity and/or the match score.
3. The method according to claim 2, wherein the step of inputting the treatment thread field and the standard treatment thread field into a semantic ordering model to obtain a matching score between the treatment thread field and the standard treatment thread field output by the semantic ordering model comprises:
determining a plurality of standard treatment threads related to the treatment thread field based on the semantic similarity;
respectively inputting the plurality of standard treatment thread fields and the treatment thread fields into the semantic sorting model to obtain a plurality of matching scores output by the semantic sorting model;
the determining a standard treatment thread matching the treatment thread field based on the semantic similarity and/or the match score comprises:
and sequencing the plurality of matching scores from high to low, and determining the standard treatment thread corresponding to the matching score which is sequenced first and is greater than or equal to a preset threshold as the standard treatment thread matched with the treatment thread field.
4. The therapy thread mining method of any one of claims 1 to 3, wherein the determining text related to a therapy thread comprises:
determining a target medical text comprising at least one of a clinical trial result class text, a treatment class guideline text, and a treatment class clinical trial text;
and carrying out treatment thread relevance classification on the target medical text to obtain a text related to a treatment thread.
5. A retrieval method, comprising:
receiving a target retrieval keyword sent by a user terminal, wherein the target retrieval keyword comprises a target treatment thread and/or a target medicine;
determining medical data associated with the target retrieval keyword from predetermined medical data of each medicine under each treatment thread;
each treatment thread to which the medical data corresponds is determined based on the treatment thread mining method of any one of claims 1 to 4, the medical data comprising at least one of guideline data, clinical trial data, and epidemiological data.
6. The searching method according to claim 5, wherein in a case where the target search keyword includes the target drug, the determining, from the predetermined medical data of each drug under each treatment thread, the medical data associated with the target search keyword includes:
and determining clinical test data of the indications corresponding to the target medicines under each treatment thread from predetermined medical data of the medicines under each treatment thread, wherein the clinical test data comprises at least one of a medicine research and development stage, a highest research and development stage of the indications corresponding to the target medicines and an approval mechanism for marketing the medicines.
7. The searching method according to claim 5, wherein in a case where the target search keyword includes the target treatment thread, the determining, from the predetermined medical data of each drug under each treatment thread, medical data associated with the target search keyword, further comprises:
acquiring medical record data of a patient;
structuring the patient medical record data to obtain structured medical record data;
and monitoring the patient medical record data based on the structured medical record data and the medical record medical quality control index threshold value under the target treatment thread.
8. The retrieval method of claim 7, wherein the monitoring the patient medical record data based on the structured medical record data and medical record medical quality control indicator thresholds under the target treatment thread comprises:
and estimating adverse reaction data corresponding to the patient medical record data under the target treatment thread based on the structured data of the risk factors contained in the structured medical record data and the predetermined relationship between the adverse reaction of different drug treatments and the risk factors generating the adverse reaction.
9. A treatment thread mining device, comprising:
the text determining unit is used for determining texts related to the treatment threads;
the entity identification unit is used for carrying out treatment thread entity identification on the text and taking an entity obtained by identification as a treatment thread field;
and the treatment thread determining unit is used for determining a standard treatment thread matched with the treatment thread field based on the semantics of the treatment thread field and the semantics of a standard treatment thread field in a treatment thread dictionary, and determining the standard treatment thread matched with the treatment thread field as the treatment thread corresponding to the text.
10. A retrieval apparatus, comprising:
the system comprises a keyword receiving unit, a target searching unit and a target searching unit, wherein the keyword receiving unit is used for receiving a target searching keyword sent by a user terminal, and the target searching keyword comprises a target treatment thread and/or a target medicine;
the medical data determining unit is used for determining medical data associated with the target retrieval keyword from the predetermined medical data of each medicine under each treatment thread; each treatment thread in the medical data is determined based on the treatment thread mining method of any one of claims 1 to 4, the medical data comprising at least one of guideline data, clinical trial data, and epidemiological data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210343899.1A CN114822859B (en) | 2022-03-31 | 2022-03-31 | Treatment thread mining and searching method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210343899.1A CN114822859B (en) | 2022-03-31 | 2022-03-31 | Treatment thread mining and searching method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114822859A true CN114822859A (en) | 2022-07-29 |
| CN114822859B CN114822859B (en) | 2023-11-03 |
Family
ID=82533246
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210343899.1A Active CN114822859B (en) | 2022-03-31 | 2022-03-31 | Treatment thread mining and searching method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114822859B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180068406A1 (en) * | 2016-09-07 | 2018-03-08 | Ucb Biopharma Sprl | Method of generating, storing and mining data related to key opinion leaders in scientific fields and computer system configured for presenting an explorable graphical user interface |
| KR101875306B1 (en) * | 2017-01-11 | 2018-07-05 | 전북대학교산학협력단 | System for providing disease information using cluster of medicine teminologies |
| CN109670177A (en) * | 2018-12-20 | 2019-04-23 | 翼健(上海)信息科技有限公司 | One kind realizing the semantic normalized control method of medicine and control device based on LSTM |
| CN113539515A (en) * | 2021-07-01 | 2021-10-22 | 上海药慧信息技术有限公司 | Clinical demand mining method and device, electronic equipment and storage medium |
| WO2021232591A1 (en) * | 2020-05-22 | 2021-11-25 | 平安国际智慧城市科技股份有限公司 | User data classification method and apparatus based on internet of things data, computer device and storage medium |
| CN113889279A (en) * | 2021-09-28 | 2022-01-04 | 北京华彬立成科技有限公司 | Combination therapy information mining and inquiring method, device and electronic equipment |
| CN113889281A (en) * | 2021-11-17 | 2022-01-04 | 重庆邮电大学 | Chinese medical intelligent entity identification method and device and computer equipment |
-
2022
- 2022-03-31 CN CN202210343899.1A patent/CN114822859B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180068406A1 (en) * | 2016-09-07 | 2018-03-08 | Ucb Biopharma Sprl | Method of generating, storing and mining data related to key opinion leaders in scientific fields and computer system configured for presenting an explorable graphical user interface |
| KR101875306B1 (en) * | 2017-01-11 | 2018-07-05 | 전북대학교산학협력단 | System for providing disease information using cluster of medicine teminologies |
| CN109670177A (en) * | 2018-12-20 | 2019-04-23 | 翼健(上海)信息科技有限公司 | One kind realizing the semantic normalized control method of medicine and control device based on LSTM |
| WO2021232591A1 (en) * | 2020-05-22 | 2021-11-25 | 平安国际智慧城市科技股份有限公司 | User data classification method and apparatus based on internet of things data, computer device and storage medium |
| CN113539515A (en) * | 2021-07-01 | 2021-10-22 | 上海药慧信息技术有限公司 | Clinical demand mining method and device, electronic equipment and storage medium |
| CN113889279A (en) * | 2021-09-28 | 2022-01-04 | 北京华彬立成科技有限公司 | Combination therapy information mining and inquiring method, device and electronic equipment |
| CN113889281A (en) * | 2021-11-17 | 2022-01-04 | 重庆邮电大学 | Chinese medical intelligent entity identification method and device and computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114822859B (en) | 2023-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kreimeyer et al. | Natural language processing systems for capturing and standardizing unstructured clinical information: a systematic review | |
| CN111986770B (en) | Prescription medication auditing method, device, equipment and storage medium | |
| CN111274806B (en) | Method and device for recognizing word segmentation and part of speech and method and device for analyzing electronic medical record | |
| US8600772B2 (en) | Systems and methods for interfacing with healthcare organization coding system | |
| Tashkandi et al. | Efficient in-database patient similarity analysis for personalized medical decision support systems | |
| CN111465990B (en) | Method and system for clinical trials of healthcare | |
| CN113345577B (en) | Diagnosis and treatment auxiliary information generation method, model training method, device, equipment and storage medium | |
| CA2705175C (en) | Systems and methods for generating subsets of electronic healthcare-related documents | |
| CN113539515B (en) | Clinical demand mining method and device, electronic equipment and storage medium | |
| CN112071431B (en) | Clinical path automatic generation method and system based on deep learning and knowledge graph | |
| CN112699230A (en) | Malignant tumor diagnosis and treatment knowledge acquisition method and device | |
| CN114822859B (en) | Treatment thread mining and searching method and device | |
| US10586616B2 (en) | Systems and methods for generating subsets of electronic healthcare-related documents | |
| CN112735584A (en) | Malignant tumor diagnosis and treatment auxiliary decision generation method and device | |
| CN109840275B (en) | Method, device and equipment for processing medical search statement | |
| CN116992839B (en) | Automatic generation method, device and equipment for medical records front page | |
| Junior et al. | A study of the influence of textual features in learning medical prior authorization | |
| US20240112765A1 (en) | Method and system for clinical trials matching | |
| Prabhakar et al. | Neural language modeling of unstructured clinical notes for automated patient phenotyping | |
| CN116895385A (en) | Target information analysis method and target information query method | |
| CN117038094A (en) | Method and system for mining and analyzing specific disease knowledge based on clinical data | |
| CN114386397A (en) | Disease code matching method and device for non-standard disease name and computer equipment | |
| CN114520047A (en) | Hospital performance accounting system, method and corresponding equipment and storage medium | |
| KR20230075780A (en) | Similarity calculation method between rare disease clinical trial documents and similarity calculation device between rare disease clinical trial documents | |
| CN116069904A (en) | Electronic medical record searching method, system, device, storage medium and product |
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 |