CN111506673A - Medical record classification code determination method and device - Google Patents
Medical record classification code determination method and device Download PDFInfo
- Publication number
- CN111506673A CN111506673A CN202010231543.XA CN202010231543A CN111506673A CN 111506673 A CN111506673 A CN 111506673A CN 202010231543 A CN202010231543 A CN 202010231543A CN 111506673 A CN111506673 A CN 111506673A
- Authority
- CN
- China
- Prior art keywords
- medical record
- disease
- determining
- candidate
- classification code
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Epidemiology (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The embodiment of the invention provides a method and a device for determining classification codes of medical records, wherein the method comprises the following steps: acquiring a medical record; determining candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model; determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name; determining a target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code. By the method for determining the classification codes of the medical records, the names of the target diseases can be determined through the medical record processing model, and the processing efficiency is high. And the target disease classification code of the medical record can be determined, and the efficiency of classifying and coding the medical record is improved.
Description
Technical Field
The present invention relates to the field of natural language processing technology, and more particularly, to a medical record classification code determination method and a medical record classification code determination device.
Background
Generally, after a patient visits a hospital, a doctor may fill out a medical record for the patient, which may be used to record the doctor's diagnosis of the patient's disease condition. In order to indicate the disease condition of the patient in a standardized manner, ICD-10(international Classification of diseases) Classification codes can be adopted to classify medical records, which is convenient for statistics, improvement of medical quality management, insurance cost accounting and other needs.
However, for ICD-10 classification codes, factors that affect disease classification may include etiology, pathology, location, clinical presentation, and the like. Similar diseases may have different ICD-10 classification codes, resulting in some difficulty in correctly classifying the case. In addition, doctors can express the state of illness of patients by means of the similar meaning words and short names of disease names based on their own habits when filling in the medical records of patients. Resulting in a further increase in the difficulty of classifying medical records. Generally, the medical records can be classified manually, however, the classification of the medical records by ICD-10 is complicated, so the manual classification is inefficient.
Disclosure of Invention
In view of the above problems, embodiments of the present invention are provided to provide a medical record classification code determination method and a corresponding medical record classification code determination apparatus that overcome or at least partially solve the above problems.
In order to solve the above problems, an embodiment of the present invention discloses a method for determining a classification code of a medical record, including:
acquiring a medical record;
determining candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model;
determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name;
determining a target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code.
Optionally, the step of determining, by using a preset medical record processing model, candidate disease names corresponding to the medical record and a similarity ranking corresponding to the candidate disease names includes:
determining whether a disease name matched with a preset disease classification code table exists in the medical record or not;
if a disease name matched with a preset disease classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the disease name and the disease classification code table;
if the disease name matched with a preset disease classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
Optionally, the step of determining, by using a preset medical record processing model, candidate disease names corresponding to the medical record and a similarity ranking corresponding to the candidate disease names includes:
determining whether a synonymous name matched with a preset synonymous classification code table exists in the medical record or not;
if the synonymy name matched with a preset synonymy classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the synonymy name and the synonymy classification code table;
if the synonymy name matched with a preset synonymy classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
Optionally, the medical record treatment model comprises a first treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
determining the reverse file frequency of each case character in the case record by adopting a first processing model and a preset reverse order dictionary table; the reverse order dictionary table comprises a mapping relation between each classification code table character in the disease classification code table and the occurrence frequency corresponding to the disease character;
determining at least one candidate disease name in the disease classification code table by adopting the medical record characters;
determining the similarity between the medical record and the candidate disease name by adopting the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
Optionally, the medical record treatment model comprises a second treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
dividing the medical record into at least one word segmentation by adopting a second processing model and a preset medical dictionary;
determining at least one candidate disease name corresponding to the segmented word by adopting the medical dictionary;
determining similarity between the candidate disease name and the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
Optionally, the medical record treatment model comprises a third treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
acquiring a medical record sentence vector of the medical record by adopting a third processing model;
determining the similarity between the medical record sentence vector and a preset disease sentence vector of the disease name;
determining at least one candidate disease name among the disease names using the similarity, and determining a similarity ranking of the candidate disease names.
Optionally, the step of determining the target disease name corresponding to the medical record by using the candidate disease name and the similarity ranking corresponding to the candidate disease name includes:
determining the occurrence frequency of the candidate disease names based on the candidate disease names respectively corresponding to at least two medical record processing models;
and determining the target disease name corresponding to the medical record by adopting the occurrence frequency of the candidate disease names and the similarity ranking of the candidate disease names.
The embodiment of the invention also discloses a medical record classification code determining device, which comprises:
the acquisition module is used for acquiring medical record records;
the ordering determination module is used for determining candidate disease names corresponding to the medical record and similarity ordering corresponding to the candidate disease names by adopting a preset medical record processing model;
the name determining module is used for determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity sequence corresponding to the candidate disease name;
the classification code determining module is used for determining a target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code.
Optionally, the rank determining module includes:
determining whether a disease name matched with a preset disease classification code table exists in the medical record or not;
if a disease name matched with a preset disease classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the disease name and the disease classification code table;
if the disease name matched with a preset disease classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
Optionally, the rank determining module includes:
determining whether a synonymous name matched with a preset synonymous classification code table exists in the medical record or not;
if the synonymy name matched with a preset synonymy classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the synonymy name and the synonymy classification code table;
if the synonymy name matched with a preset synonymy classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
Optionally, the medical record treatment model comprises a first treatment model;
the rank determination module includes:
determining the reverse file frequency of each case character in the case record by adopting a first processing model and a preset reverse order dictionary table; the reverse order dictionary table comprises a mapping relation between each classification code table character in the disease classification code table and the occurrence frequency corresponding to the disease character;
determining at least one candidate disease name in the disease classification code table by adopting the medical record characters;
determining the similarity between the medical record and the candidate disease name by adopting the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
Optionally, the medical record treatment model comprises a second treatment model;
the rank determination module includes:
dividing the medical record into at least one word segmentation by adopting a second processing model and a preset medical dictionary;
determining at least one candidate disease name corresponding to the segmented word by adopting the medical dictionary;
determining similarity between the candidate disease name and the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
Optionally, the medical record treatment model comprises a third treatment model;
the rank determination module includes:
acquiring a medical record sentence vector of the medical record by adopting a third processing model;
determining the similarity between the medical record sentence vector and a preset disease sentence vector of the disease name;
determining at least one candidate disease name among the disease names using the similarity, and determining a similarity ranking of the candidate disease names.
Optionally, the name determining module includes:
determining the occurrence frequency of the candidate disease names based on the candidate disease names respectively corresponding to at least two medical record processing models;
and determining the target disease name corresponding to the medical record by adopting the occurrence frequency of the candidate disease names and the similarity ranking of the candidate disease names.
The embodiment of the invention also discloses a device, which comprises:
one or more processors; and
one or more machine-readable media having instructions stored thereon, which when executed by the one or more processors, cause the apparatus to perform one or more methods as described in embodiments of the invention.
Embodiments of the invention also disclose one or more machine-readable media having instructions stored thereon, which when executed by one or more processors, cause the processors to perform one or more methods as described in embodiments of the invention.
The embodiment of the invention has the following advantages:
according to the classification code determination method for the medical records, the medical record is processed by adopting a preset medical record processing model, and candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names are determined; and determining the target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name, so that the target disease name can be determined through a medical record processing model, and the processing efficiency is high. And determining the target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name, so that the target disease classification code of the medical record can be determined, and the efficiency of classifying the medical record is improved.
Drawings
FIG. 1 is a flowchart illustrating the steps of a method for determining classification codes of medical records according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating the steps of another method for determining classification codes of medical records according to an embodiment of the present invention;
FIG. 3 is a block diagram of an embodiment of a device for determining classification codes of medical records according to the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Referring to fig. 1, a flowchart illustrating steps of an embodiment of a method for determining a classification code of a medical record according to an embodiment of the present invention is shown, which may specifically include the following steps:
in an embodiment of the present invention, at least one medical record may be obtained. The medical record can be used to record the patient's condition. The medical record can be filled in by a doctor in a preset medical system, or the medical record can be converted into an electronic record by an optical character recognition mode after the medical record is filled in by the doctor, so that the medical record of an electronic version can be obtained.
in an embodiment of the present invention, the medical record processing model may process the medical record, convert the medical record into other expression manners, such as vectors, word segmentation, and the like, so as to match the medical record with a disease name expressed by a preset specification, and compare similarities between the medical record and the disease name, thereby determining at least one candidate disease name corresponding to the medical record and a similarity rank corresponding to the at least one candidate disease name.
In an embodiment of the present invention, the similarity ranking may be a ranking of similarities between the medical record and the candidate disease names. The higher the similarity ranking, the higher the similarity between the medical record and the candidate disease name may be.
In an embodiment of the present invention, the medical record processing model may have at least one, and each of the medical record processing models may determine a set of candidate disease names and similarity ranks corresponding to the candidate disease names. Each set of the candidate disease names and the similarity rankings corresponding to the candidate disease names may include at least one candidate disease name and a similarity ranking corresponding to the at least one candidate disease name. Therefore, when the medical record is processed by adopting a plurality of medical record processing models, a plurality of groups of candidate disease names and similarity ranks corresponding to the candidate disease names can be obtained. And more accurate target disease names can be obtained based on a plurality of medical record processing models.
103, determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity sequence corresponding to the candidate disease name;
in the embodiment of the present invention, the candidate disease names and the similarity ranks corresponding to the candidate disease names may be adopted, and at least one candidate disease name with a higher similarity rank is selected as the target disease name corresponding to the medical record.
In the embodiment of the present invention, in the case of having multiple medical record models, at least one candidate disease name with a higher similarity ranking in each group may be determined as a target disease name corresponding to the medical record based on multiple groups of candidate disease names and similarity rankings corresponding to the candidate disease names.
In the embodiment of the present invention, in the case where there are a plurality of medical record models, the medical record models may further have weights. The target disease name of the medical record can be determined preferentially based on the candidate disease name determined by the medical record model with higher weight and the similarity ranking corresponding to the candidate disease name.
In embodiments of the present invention, the medical record may have one target disease name or more than two target disease names, depending on the actual condition of the patient. The target disease name may simultaneously express one disease diagnosis, or two or more disease diagnoses. For example, the disease name "subarachnoid hemorrhage" may be one disease diagnosis, and the disease name "carotid aneurysm rupture with subarachnoid hemorrhage" may be two disease diagnoses. Thus, a disease name expressing diagnosis of two or more diseases among the candidate disease names can be selected as a target disease name according to actual needs. In cases where a patient cannot be expressed with one disease name, for example, the cause of the patient is expressed with one disease name and the clinical manifestation of the patient is expressed with another disease name, or sequela is expressed with a second disease name, two or more candidate disease names may be used as target disease names.
In the embodiment of the invention, after at least two disease names are determined, the most appropriate target disease name can be selected from the target disease names through manual determination, so that the target disease name can be selected manually through machine assistance, and the processing efficiency of manual classification is improved.
In an embodiment of the present invention, the disease classification code table may include a mapping relationship between a disease name and a disease classification code. Therefore, after at least one target disease name is determined, the disease classification code table can be adopted to search the disease classification code corresponding to the target disease name in the disease classification code table, so that at least one target disease classification code corresponding to the medical record can be determined.
According to the classification code determination method for the medical records, the medical record is processed by adopting a preset medical record processing model, and candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names are determined; and determining the target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name, so that the target disease name can be determined through a medical record processing model, and the processing efficiency is high. And determining the target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name, so that the target disease classification code of the medical record can be determined, and the efficiency of classifying the medical record is improved.
Referring to fig. 2, a flowchart illustrating steps of another embodiment of a method for determining a classification code of a medical record according to an embodiment of the present invention is shown, which may specifically include the following steps:
in an embodiment of the present invention, at least one medical record may be obtained. The medical record can be used to record the patient's condition. The medical record can be filled in by a doctor in a preset medical system, or the medical record can be converted into an electronic record by an optical character recognition mode after the medical record is filled in by the doctor, so that the medical record of an electronic version can be obtained.
in an embodiment of the present invention, the medical record processing model may process the medical record, convert the medical record into other expression manners, such as vectors, word segmentation, and the like, so as to match the medical record with a disease name expressed by a preset specification, and compare similarities between the medical record and the disease name, thereby determining at least one candidate disease name corresponding to the medical record and a similarity rank corresponding to the at least one candidate disease name.
In an embodiment of the present invention, the similarity ranking may be a ranking of similarities between the medical record and the candidate disease names. The higher the similarity ranking, the higher the similarity between the medical record and the candidate disease name may be.
In an embodiment of the present invention, the medical record processing model may have at least one, and each of the medical record processing models may determine a set of candidate disease names and similarity ranks corresponding to the candidate disease names. Each set of the candidate disease names and the similarity rankings corresponding to the candidate disease names may include at least one candidate disease name and a similarity ranking corresponding to the at least one candidate disease name. Therefore, when the medical record is processed by adopting a plurality of medical record processing models, a plurality of groups of candidate disease names and similarity ranks corresponding to the candidate disease names can be obtained. And more accurate target disease names can be obtained based on a plurality of medical record processing models.
In an embodiment of the present invention, the step of determining the candidate disease names corresponding to the medical record and the similarity ranks corresponding to the candidate disease names by using a preset medical record processing model includes:
s11, determining whether a disease name matched with a preset disease classification code table exists in the medical record or not;
in the embodiment of the present invention, the disease classification code table may store a disease name. The medical record may include the name of the disease. At this time, the target disease name corresponding to the medical record does not need to be determined by adopting a preset medical record processing model, and whether the disease name matched with a preset disease classification code table exists in the medical record can be determined directly by matching keywords.
In the embodiment of the invention, a disease diagnosis column is arranged in the medical record, and the disease diagnosis information of a doctor on a patient can be filled in the disease diagnosis column, so that whether a disease name matched with a preset disease classification code table exists in the medical record can be determined in the disease diagnosis column, and the efficiency of determining a target disease classification code is further improved.
S12, if a disease name matched with a preset disease classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by using the disease name and the disease classification code table;
in the embodiment of the present invention, if a disease name matching a preset disease classification code table exists in the medical record, it may be considered that a patient fills in a standardized disease diagnosis at this time, and a preset medical record processing model may not be used to determine a target disease name corresponding to the medical record, but a disease classification code corresponding to the disease name may be used as a target disease classification code based on a mapping relationship between the disease name and the disease classification code in the disease classification code table, so that the target disease classification code corresponding to the medical record may be determined efficiently, and classification of the medical record may be implemented.
And S13, if the disease name matched with a preset disease classification code table does not exist in the medical record, determining the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name by adopting a preset medical record processing model.
In the embodiment of the present invention, if there is no disease name matching the preset disease classification code table in the medical record, a preset medical record processing model may be adopted to process the medical record, determine a candidate disease name corresponding to the medical record and a similarity ranking corresponding to the candidate disease name, and determine at least one target disease name of the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name, so as to obtain a target disease classification code corresponding to the medical record.
In an embodiment of the present invention, the step of determining the candidate disease names corresponding to the medical record and the similarity ranks corresponding to the candidate disease names by using a preset medical record processing model includes:
s21, determining whether a synonymous name matched with a preset synonymous classification code table exists in the medical record;
in the embodiment of the present invention, a synonymous classifier table may be set based on the disease classifier table. The synonym classification code table may include a mapping relationship of synonyms of the disease names and the disease classification codes.
In an embodiment of the present invention, the synonymous name may be a name stored in the synonymous classifier table and having the same meaning as the disease name. The medical records may include the synonymous name. At this time, the target disease name corresponding to the medical record does not need to be determined by adopting a preset medical record processing model, and whether the synonymous name matched with the preset synonymous classification code table exists in the medical record can be determined directly in a way of matching the keyword.
In the embodiment of the invention, a disease diagnosis column is arranged in the medical record, and the disease diagnosis information of a doctor on a patient can be filled in the disease diagnosis column, so that whether a synonymous name matched with a preset synonymous classification code table exists in the medical record can be determined in the disease diagnosis column, and the efficiency of determining a target disease classification code is further improved.
S22, if the medical record has the synonymous name matched with the preset synonymous classified code table, determining the target disease classified code corresponding to the medical record by adopting the synonymous name and the synonymous classified code table;
in the embodiment of the present invention, if the synonymous name matching the preset synonymous classification code table exists in the medical record, the target disease name corresponding to the medical record may not be determined by using the preset medical record processing model, but the disease classification code corresponding to the synonymous name may be used as the target disease classification code based on the mapping relationship between the synonymous name and the disease classification code in the synonymous classification code table, so that the target disease classification code corresponding to the medical record may be determined efficiently, and the classification code of the medical record may be implemented.
And S23, if the medical record does not have the synonymous name matched with the preset synonymous classification code table, determining the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name by adopting a preset medical record processing model.
In the embodiment of the present invention, if there is no synonymous name matching a preset synonymous classification code table in the medical record, a preset medical record processing model may be adopted to process the medical record, determine a candidate disease name corresponding to the medical record and a similarity ranking corresponding to the candidate disease name, and determine at least one target disease name of the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name, so as to obtain a target disease classification code corresponding to the medical record.
In one embodiment of the invention, the medical record processing model comprises a first processing model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
s31, determining the reverse file frequency of each case character in the case record by adopting a first processing model and a preset reverse order dictionary table; the reverse order dictionary table comprises a mapping relation between each classification code table character in the disease classification code table and the occurrence frequency corresponding to the classification code table character;
in the embodiment of the present invention, a plurality of classification code table characters can be extracted from the disease classification code table. The classification codebook characters may be characters included in the disease classification codebook. The classification code table characters may correspond to a frequency of occurrence. The frequency of occurrence may be the number of disease names in the disease classification code, including the characters of the classification code table. The reverse order dictionary table may be used to represent a mapping relationship between the characters of the classification code table and the occurrence frequency corresponding to the classification code table.
In the embodiment of the present invention, the medical record characters may be characters included in the medical record. The first processing model and the reverse order dictionary table may be used to determine a reverse file frequency for each case character in the case record.
In the embodiment of the present invention, in a case where the occurrence frequency of the classification code table characters is low, if there is a medical record character that is the same as the classification code table character with the low occurrence frequency in the medical record, it may be considered that the medical record may have a higher correlation with a disease name including the classification code table character. From this, the inverse file frequency of the case characters can be determined. The inverse file frequency may be used to give higher weight to case characters that are the same as the less frequently occurring taxonomic charater characters.
In the embodiment of the invention, the reverse file frequency can be obtained by adopting a first processing model and a reverse order dictionary table. Specifically, the first process model may be represented by the following function:
IDFi＝1/log2(n+1)
wherein, IDFiThe frequency of the reverse file of the characters of the medical record is n, and the occurrence frequency corresponding to the characters of the classification code table which are the same as the characters of the medical record is n. Can adoptAnd determining the occurrence frequency of the characters of the classification code table which are the same as the characters of the medical case by the reverse order dictionary table.
S32, determining at least one candidate disease name in the disease classification code table by adopting the medical record characters;
in the embodiment of the present invention, the disease name including at least one of the medical record characters in the disease classification code table may be used as a candidate disease name. Whereby at least one candidate disease name may be determined in said disease class code table.
S33, determining the similarity between the medical record and the candidate disease name by adopting the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record;
in the embodiment of the present invention, the medical record may have at least one medical record character that is the same as a character included in a candidate disease name. The similarity between the medical record and the candidate disease name can be determined by using the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record.
Specifically, in the case record, the greater the number of case characters identical to the characters included in the candidate disease name is, and the higher the value of the reverse file frequency of the case characters is, the higher the degree of similarity between the case record and the candidate disease name is.
As an alternative embodiment of the present invention, the similarity may be a sum of squares of inverse file frequencies of at least one medical record character in the medical record that is the same as the character contained in the candidate disease name.
In an embodiment of the present invention, after determining at least one candidate disease name, a minimum edit distance between the candidate disease name and the medical record may also be determined. And in the case that a disease diagnosis column is arranged in the medical record, and the disease diagnosis column can be filled with the disease diagnosis information of the patient by the doctor, the candidate disease name and the minimum editing distance between the candidate disease name and the disease diagnosis information can be determined.
In the embodiment of the present invention, the similarity between the medical record and the candidate disease name may be determined by integrating the inverse file frequency of the inverse file frequency and the minimum editing distance. Specifically, the similarity may be the minimum edit distance and a weighted average between the square sum of the inverse document frequencies of at least one medical record character identical to the character included in the candidate disease name in the medical record.
And S34, sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
In the embodiment of the present invention, the similarity may be adopted, and the candidate disease names may be ranked from high similarity to low similarity, so that the similarity ranking of the candidate disease names may be obtained. Therefore, the candidate disease names corresponding to the medical record records and the similarity ranking of the candidate disease names can be determined based on the characters contained in the medical record records, and under the condition that the similarity between the medical record records and the disease names is high, good accuracy can be obtained at a high speed.
In one embodiment of the invention, the medical record processing model comprises a second processing model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
s41, dividing the medical record into at least one word segmentation by adopting a second processing model and a preset medical dictionary;
in an embodiment of the invention, the medical dictionary may include proprietary words of several existing medical categories. The second processing model may be a chinese word segmentation model. The second processing model may employ the medical dictionary to divide the medical records into at least one segmentation.
As an alternative embodiment of the present invention, the second processing model may be a statistical-based chinese word segmentation model. Such as hidden markov models, conditional random field models, etc., as the present invention is not limited in this respect. The medical record may be segmented using the second processing model to obtain at least one segmented combination, and then the medical dictionary may be used to correct the at least one segmented combination, for example, to re-combine the proper vocabulary segmented into several segmented words, or to adjust the segmentation position in the medical record to correct the segmentation result of the chinese segmentation model.
As an alternative embodiment of the present invention, the second processing model may be a neural network-based segmentation model, such as L STM (L ong Short-Term Memory) model, FCN (full volume network) model, etc. the second processing model may be trained using the medical dictionary, and then the medical record may be segmented using the trained second processing model to divide the medical record into at least one segmentation.
S42, determining at least one candidate disease name corresponding to the participle by adopting the medical dictionary;
in an embodiment of the present invention, the medical dictionary may further include links between specialized vocabularies. Specifically, the relation between the professional vocabularies can comprise a near-synonym relation, a synonym relation and an upper-lower-level word relation. The specialized vocabulary may also include the names of diseases in the disease category code table.
In an embodiment of the present invention, the medical dictionary may be used to determine, for each of the segmented words, at least one professional vocabulary associated with the segmented word. Thereafter, at least one candidate disease name may be determined using the segmentation and the specialized vocabulary. Specifically, the disease classification code table includes the word or the disease name of the professional vocabulary associated with the word as the candidate disease name.
Optionally, the professional vocabulary associated with the word segmentation may be the disease name included in the disease classification code table, and the disease name associated with the word segmentation may be used as the candidate disease name.
S43, determining the similarity between the candidate disease name and the medical record;
in the embodiment of the present invention, the candidate disease name and the medical record may be compared to determine the similarity therebetween. Specifically, the similarity between the medical record and the candidate disease name may be determined by using a reverse file frequency of medical record characters in the medical record, which are the same as characters included in the candidate disease name. The minimum edit distance between the medical record and the candidate disease name can be determined by adopting the minimum edit distance between the medical record and the candidate disease name, and the similarity between the medical record and the candidate disease name can be determined by adopting a word vector. The invention is not limited in this regard.
And S44, sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
In the embodiment of the present invention, the similarity may be adopted, and the candidate disease names may be ranked from high similarity to low similarity, so that the similarity ranking of the candidate disease names may be obtained. Therefore, the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names can be determined based on the words contained in the medical record. In a case that the medical record is concise, the candidate disease names and the similarity ranks corresponding to the candidate disease names may be obtained based on the medical dictionary. With the completion of the medical dictionary, the word segmentation accuracy of the second processing model can be further improved.
In one embodiment of the invention, the medical record processing model comprises a third processing model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
s51, acquiring a medical record sentence vector of the medical record by adopting a third processing model;
in an embodiment of the present invention, the third processing model may be used to convert a segment of a string in the medical field into a vector form of expression. The third processing model may be Word2vec, BERT (Bidirectional encoding representation from transforms), and the like, which may convert a segment of a string into a vector representation. The third processing model may be trained using corpora of the medical field, such as a disease classification code table, a medical dictionary, a medical textbook, a medical article, and the like, so that a third processing model suitable for the medical field may be obtained.
In the embodiment of the present invention, the medical record may be input into the third processing model, so that a vector-form expression of the medical record processing model, that is, a medical sentence vector may be obtained. As an example of the present invention, the third processing model may be a BERT model, the medical records may be input into the third processing model, and the third processing model may output a 128-dimensional medical sentence vector.
S52, determining the similarity between the medical case sentence vector and a preset disease sentence vector of the disease name;
in an embodiment of the present invention, the disease sentence vector may be a vector-form expression of the disease name. The third processing model may be adopted in advance to convert the disease names in the disease classification code table into the disease sentence vectors.
In the embodiment of the present invention, the medical case sentence vector and the disease sentence vector may be compared to determine the similarity therebetween. Specifically, the cosine distance between the medical record sentence vector and the disease sentence vector may be calculated, so that the similarity between the medical record sentence vector and the disease sentence vector is determined by using the cosine distance. The closer the cosine distance is to 1, the higher the similarity between the case sentence vector and the disease sentence vector can be considered.
S53, determining at least one candidate disease name in the disease names by using the similarity, and determining similarity ranking of the candidate disease names.
In the embodiment of the present invention, the similarity may be adopted, and the candidate disease names may be ranked from high similarity to low similarity, so that the similarity ranking of the candidate disease names may be obtained. Therefore, the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names can be determined based on the medical record. Under the condition that the medical record is more complex, the cosine distance between sentences can be calculated based on the third processing model, and the candidate disease names and the similarity ranking corresponding to the candidate disease names are obtained. With the increase of the number of medical records processed by the third processing model, the third processing model can be further trained by adopting the medical records, so that the accuracy of the third processing model is further improved.
in an embodiment of the present invention, each of the medical record processing models may determine a set of candidate disease names and similarity ranks corresponding to the candidate disease names. Each set of the candidate disease names and the similarity rankings corresponding to the candidate disease names may include at least one candidate disease name and a similarity ranking corresponding to the at least one candidate disease name.
In the embodiment of the present invention, at least two medical record processing models may be used to process the medical record, so that at least two sets of candidate disease names and similarity ranks corresponding to the candidate disease names may be obtained. The candidate disease names in each group may be different, and the similarity ranking corresponding to the candidate disease names may also be different. Thereby, the frequency of each candidate disease name in the at least two groups of candidate disease names output by the at least two case processing models and the similarity ranking corresponding to the candidate disease names can be determined.
204, determining a target disease name corresponding to the medical record by adopting the occurrence frequency of the candidate disease names and the similarity ranking of the candidate disease names;
in the embodiment of the present invention, the target disease name corresponding to the medical record may be determined by using the frequency of occurrence of the candidate disease names and the similarity ranking of the candidate disease names. Specifically, the candidate disease names with a high frequency of occurrence and a high similarity ranking of the candidate disease names may be used as the target disease names corresponding to the medical record.
As an optional implementation manner of the present invention, in at least one group of candidate disease names and similarity ranks corresponding to the disease names, candidate disease names with similarity ranks larger than a preset rank may be selected; and then at least one candidate disease name with the frequency greater than a preset threshold in the candidate disease names with the similarity ranking greater than the preset ranking is taken as the target disease name.
As an alternative embodiment of the present invention, candidate disease names with occurrence frequency greater than a preset threshold may be selected, and then at least one candidate disease name with higher similarity ranking in a higher-weight case processing model is selected as a target disease name based on the weight of the case processing model and the similarity ranking of the candidate disease names.
In an embodiment of the present invention, the disease classification code table may include a mapping relationship between a disease name and a disease classification code. Therefore, after at least one target disease name is determined, the disease classification code table can be adopted to search the disease classification code corresponding to the target disease name in the disease classification code table, so that at least one target disease classification code corresponding to the medical record can be determined.
According to the classification code determination method for the medical records, the medical record is processed by adopting a preset medical record processing model, and candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names are determined; and determining the target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name, so that the target disease name can be determined through a medical record processing model, and the processing efficiency is high. And determining the target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name, so that the target disease classification code of the medical record can be determined, and the efficiency of classifying the medical record is improved.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the illustrated order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments of the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the invention.
Referring to fig. 3, a block diagram of a structure of an embodiment of the device for determining classification codes of medical records of the present invention is shown, which may specifically include the following modules:
an obtaining module 301, configured to obtain a medical record;
a rank determination module 302, configured to determine, by using a preset medical record processing model, a candidate disease name corresponding to the medical record and a similarity rank corresponding to the candidate disease name;
a name determining module 303, configured to determine a target disease name corresponding to the medical record by using the candidate disease name and the similarity ranking corresponding to the candidate disease name;
a classification code determining module 304, configured to determine a target disease classification code corresponding to the medical record by using a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code.
In one embodiment of the present invention, the rank determining module includes:
determining whether a disease name matched with a preset disease classification code table exists in the medical record or not;
if a disease name matched with a preset disease classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the disease name and the disease classification code table;
if the disease name matched with a preset disease classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
In one embodiment of the present invention, the rank determining module includes:
determining whether a synonymous name matched with a preset synonymous classification code table exists in the medical record or not;
if the synonymy name matched with a preset synonymy classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the synonymy name and the synonymy classification code table;
if the synonymy name matched with a preset synonymy classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
In one embodiment of the present invention, the medical record processing model comprises a first processing model;
the rank determination module includes:
determining the reverse file frequency of each case character in the case record by adopting a first processing model and a preset reverse order dictionary table; the reverse order dictionary table comprises a mapping relation between each classification code table character in the disease classification code table and the occurrence frequency corresponding to the disease character;
determining at least one candidate disease name in the disease classification code table by adopting the medical record characters;
determining the similarity between the medical record and the candidate disease name by adopting the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
In one embodiment of the present invention, the medical record processing model comprises a second processing model;
the rank determination module includes:
dividing the medical record into at least one word segmentation by adopting a second processing model and a preset medical dictionary;
determining at least one candidate disease name corresponding to the segmented word by adopting the medical dictionary;
determining similarity between the candidate disease name and the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
In one embodiment of the present invention, the medical record processing model comprises a third processing model;
the rank determination module includes:
acquiring a medical record sentence vector of the medical record by adopting a third processing model;
determining the similarity between the medical record sentence vector and a preset disease sentence vector of the disease name;
determining at least one candidate disease name among the disease names using the similarity, and determining a similarity ranking of the candidate disease names.
In an embodiment of the present invention, the name determining module includes:
determining the occurrence frequency of the candidate disease names based on the candidate disease names respectively corresponding to at least two medical record processing models;
and determining the target disease name corresponding to the medical record by adopting the occurrence frequency of the candidate disease names and the similarity ranking of the candidate disease names.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
An embodiment of the present invention further provides an apparatus, including:
one or more processors; and
one or more machine-readable media having instructions stored thereon, which when executed by the one or more processors, cause the apparatus to perform methods as described in embodiments of the invention.
Embodiments of the invention also provide one or more machine-readable media having instructions stored thereon, which when executed by one or more processors, cause the processors to perform the methods described in embodiments of the invention.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (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 terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, 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 terminal 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 terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
The method for determining the classification code of the medical record and the device for determining the classification code of the medical record provided by the invention are introduced in detail, and specific examples are applied in the text to explain the principle and the implementation mode of the invention, and the description of the above embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A method for determining classification codes of medical records is characterized by comprising the following steps:
acquiring a medical record;
determining candidate disease names corresponding to the medical record and similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model;
determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity ranking corresponding to the candidate disease name;
determining a target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code.
2. The method according to claim 1, wherein the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by using a preset medical record processing model comprises:
determining whether a disease name matched with a preset disease classification code table exists in the medical record or not;
if a disease name matched with a preset disease classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the disease name and the disease classification code table;
if the disease name matched with a preset disease classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
3. The method according to claim 1 or 2, wherein the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by using a preset medical treatment model comprises:
determining whether a synonymous name matched with a preset synonymous classification code table exists in the medical record or not;
if the synonymy name matched with a preset synonymy classification code table exists in the medical record, determining a target disease classification code corresponding to the medical record by adopting the synonymy name and the synonymy classification code table;
if the synonymy name matched with a preset synonymy classification code table does not exist in the medical record, a preset medical record processing model is adopted to determine the candidate disease name corresponding to the medical record and the similarity ranking corresponding to the candidate disease name.
4. The method of claim 1, wherein the medical condition treatment model comprises a first treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
determining the reverse file frequency of each case character in the case record by adopting a first processing model and a preset reverse order dictionary table; the reverse order dictionary table comprises a mapping relation between each classification code table character in the disease classification code table and the occurrence frequency corresponding to the disease character;
determining at least one candidate disease name in the disease classification code table by adopting the medical record characters;
determining the similarity between the medical record and the candidate disease name by adopting the reverse file frequency of the medical record characters which are the same as the characters contained in the candidate disease name in the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
5. The method of claim 1, wherein the medical treatment model comprises a second treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
dividing the medical record into at least one word segmentation by adopting a second processing model and a preset medical dictionary;
determining at least one candidate disease name corresponding to the segmented word by adopting the medical dictionary;
determining similarity between the candidate disease name and the medical record;
and sorting the candidate disease names by adopting the similarity to obtain the similarity sorting of the candidate disease names.
6. The method of claim 1, wherein the medical condition treatment model comprises a third treatment model;
the step of determining the candidate disease names corresponding to the medical record and the similarity ranking corresponding to the candidate disease names by adopting a preset medical record processing model comprises the following steps:
acquiring a medical record sentence vector of the medical record by adopting a third processing model;
determining the similarity between the medical record sentence vector and a preset disease sentence vector of the disease name;
determining at least one candidate disease name among the disease names using the similarity, and determining a similarity ranking of the candidate disease names.
7. The method according to claim 1, wherein the step of determining the target disease name corresponding to the medical record by using the candidate disease name and the similarity ranking corresponding to the candidate disease name comprises:
determining the occurrence frequency of the candidate disease names based on the candidate disease names respectively corresponding to at least two medical record processing models;
and determining the target disease name corresponding to the medical record by adopting the occurrence frequency of the candidate disease names and the similarity ranking of the candidate disease names.
8. A medical record classification code determining device, comprising:
the acquisition module is used for acquiring medical record records;
the ordering determination module is used for determining candidate disease names corresponding to the medical record and similarity ordering corresponding to the candidate disease names by adopting a preset medical record processing model;
the name determining module is used for determining a target disease name corresponding to the medical record by adopting the candidate disease name and the similarity sequence corresponding to the candidate disease name;
the classification code determining module is used for determining a target disease classification code corresponding to the medical record by adopting a preset disease classification code table and the target disease name; the disease classification code table includes a mapping relationship between a disease name and a disease classification code.
9. An apparatus, comprising:
one or more processors; and
one or more machine-readable media having instructions stored thereon that, when executed by the one or more processors, cause the apparatus to perform the method of one or more of claims 1-7.
10. One or more machine readable media having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method of one or more of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010231543.XA CN111506673A (en) | 2020-03-27 | 2020-03-27 | Medical record classification code determination method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010231543.XA CN111506673A (en) | 2020-03-27 | 2020-03-27 | Medical record classification code determination method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111506673A true CN111506673A (en) | 2020-08-07 |
Family
ID=71877866
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010231543.XA Pending CN111506673A (en) | 2020-03-27 | 2020-03-27 | Medical record classification code determination method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111506673A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112818085A (en) * | 2021-01-28 | 2021-05-18 | 东软集团股份有限公司 | Value range data matching method and device, storage medium and electronic equipment |
| CN112926315A (en) * | 2021-05-11 | 2021-06-08 | 北京嘉和海森健康科技有限公司 | Automatic medical term standardization method and device |
| CN113312265A (en) * | 2021-06-08 | 2021-08-27 | 富途网络科技(深圳)有限公司 | Application method of test case and related product |
| CN114388085A (en) * | 2021-11-23 | 2022-04-22 | 皖南医学院第一附属医院（皖南医学院弋矶山医院） | Real-time intelligent auxiliary ICD coding method and system based on medical record |
| CN114580354A (en) * | 2022-05-05 | 2022-06-03 | 阿里巴巴达摩院(杭州)科技有限公司 | Synonym-based information encoding method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109065157A (en) * | 2018-08-01 | 2018-12-21 | 中国人民解放军第二军医大学 | A kind of Disease Diagnosis Standard coded Recommendation list determines method and system |
| CN109785959A (en) * | 2018-12-14 | 2019-05-21 | 平安医疗健康管理股份有限公司 | A kind of disease code method and apparatus |
| CN110852076A (en) * | 2019-10-12 | 2020-02-28 | 云知声智能科技股份有限公司 | Method and device for automatic disease code conversion |
-
2020
- 2020-03-27 CN CN202010231543.XA patent/CN111506673A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109065157A (en) * | 2018-08-01 | 2018-12-21 | 中国人民解放军第二军医大学 | A kind of Disease Diagnosis Standard coded Recommendation list determines method and system |
| CN109785959A (en) * | 2018-12-14 | 2019-05-21 | 平安医疗健康管理股份有限公司 | A kind of disease code method and apparatus |
| CN110852076A (en) * | 2019-10-12 | 2020-02-28 | 云知声智能科技股份有限公司 | Method and device for automatic disease code conversion |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112818085A (en) * | 2021-01-28 | 2021-05-18 | 东软集团股份有限公司 | Value range data matching method and device, storage medium and electronic equipment |
| CN112926315A (en) * | 2021-05-11 | 2021-06-08 | 北京嘉和海森健康科技有限公司 | Automatic medical term standardization method and device |
| CN113312265A (en) * | 2021-06-08 | 2021-08-27 | 富途网络科技(深圳)有限公司 | Application method of test case and related product |
| CN114388085A (en) * | 2021-11-23 | 2022-04-22 | 皖南医学院第一附属医院（皖南医学院弋矶山医院） | Real-time intelligent auxiliary ICD coding method and system based on medical record |
| CN114580354A (en) * | 2022-05-05 | 2022-06-03 | 阿里巴巴达摩院(杭州)科技有限公司 | Synonym-based information encoding method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110502621B (en) | Question answering method, question answering device, computer equipment and storage medium | |
| US11403680B2 (en) | Method, apparatus for evaluating review, device and storage medium | |
| CN106649786B (en) | Answer retrieval method and device based on deep question answering | |
| CN111506673A (en) | Medical record classification code determination method and device | |
| CN107329949B (en) | Semantic matching method and system | |
| US9348900B2 (en) | Generating an answer from multiple pipelines using clustering | |
| US9146987B2 (en) | Clustering based question set generation for training and testing of a question and answer system | |
| US9230009B2 (en) | Routing of questions to appropriately trained question and answer system pipelines using clustering | |
| CN111708873A (en) | Intelligent question answering method and device, computer equipment and storage medium | |
| CN111539197B (en) | Text matching method and device, computer system and readable storage medium | |
| CN113869044A (en) | Keyword automatic extraction method, device, equipment and storage medium | |
| US20140351228A1 (en) | Dialog system, redundant message removal method and redundant message removal program | |
| US20220405484A1 (en) | Methods for Reinforcement Document Transformer for Multimodal Conversations and Devices Thereof | |
| CN110287337A (en) | The system and method for medicine synonym is obtained based on deep learning and knowledge mapping | |
| CN109086265B (en) | Semantic training method and multi-semantic word disambiguation method in short text | |
| CN112017744A (en) | Electronic case automatic generation method, device, equipment and storage medium | |
| CN111613341A (en) | Entity linking method and device based on semantic components | |
| CN112527977B (en) | Concept extraction method, concept extraction device, electronic equipment and storage medium | |
| CN112183104B (en) | Code recommendation method, system, corresponding equipment and storage medium | |
| CN111126084B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN114692620A (en) | Text processing method and device | |
| CN113486670B (en) | Text classification method, device, equipment and storage medium based on target semantics | |
| JP4325370B2 (en) | Document-related vocabulary acquisition device and program | |
| CN117407502A (en) | Question-answer pair extraction method and device, electronic equipment and storage medium | |
| Spithourakis et al. | Clinical text prediction with numerically grounded conditional language models |
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 |