KR20220063005A - Apparatus for generating patient health summary and method thererof - Google Patents

Abstract

Disclosed are an apparatus and a method for generating a patient health summary. The apparatus includes a preprocessing unit that extracts at least one sentence from a content of clinical information included in a clinical record of a patient, and performs preprocessing for performing part-of-speech tagging on the extracted sentence. The apparatus further includes a main processing unit that extracts candidate phrases in the form of causal phrases from the sentence preprocessed by the preprocessing unit, and generates a patient health summary through semantic analysis using preset medical information. In this case, a candidate syntax in the form of the causal syntax represents a triple in the form of <noun, verb, noun> identified from the sentence preprocessed by the preprocessing unit. According to the present invention, a medical service provider can easily check a health status of the patient.

Description

APPARATUS FOR GENERATING PATIENT HEALTH SUMMARY AND METHOD THEREROF

The present invention relates to a device and method for generating a patient health summary.

Clinical records represent interactions between patients and health care providers, such as doctors, during clinical encounters. This clinical record includes reported symptoms, clinical tests, results, diagnostic tests and results, prescribed treatment, and other essential interventions complemented by family history and social and behavioral aspects.

The main purpose of clinical records is to track a patient's health status over time. Whenever a patient visits a doctor, the doctor first thoroughly checks the patient's clinical record to obtain complete details about the patient's health before making any necessary action or decisions.

As the quantity of patient-related clinical records exceeds the threshold of human expert manual curation, there is a need for a system using advanced text mining techniques of the present era, and until recently, various techniques for mining clinical records have been developed.

However, most of the existing applications focus only on some specific type of information, such as disease status, laboratory test results, or drug-related information. On the other hand, doctors need a comprehensive profile that includes symptoms, laboratory tests, diagnosed disease and treatment plan.

Therefore, there is a need for a method that can meet the expectations of doctors and provide them with a list of necessary information as a patient health profile.

SUMMARY OF THE INVENTION An object of the present invention is to provide an apparatus and method for generating a patient health summary that can easily identify a patient's health status at a glance and save a medical professional's time.

The characteristic configuration of the present invention for achieving the object of the present invention as described above and for realizing the characteristic effects of the present invention to be described later is as follows.

According to one aspect of the present invention, there is provided an apparatus for generating a patient health summary, the apparatus comprising:

A preprocessor that extracts at least one sentence from the contents of clinical information included in the clinical record for a patient, performs preprocessing for performing part-of-speech tagging on the extracted sentence, and from the sentence on which the preprocessing is performed by the preprocessor and a main processing unit for generating a patient health summary through semantic analysis using preset medical information after extracting a candidate phrase in the form of a causal phrase, wherein the candidate phrase in the form of a causal phrase is a sentence preprocessed by the preprocessor It represents a triple of the form of <noun, verb, noun> identified from

According to another aspect of the invention, there is provided a method for generating a patient health summary, the method comprising:

Extracting at least one sentence from the contents of clinical information included in the clinical record for a patient, performing pre-processing of performing part-of-speech tagging on the extracted sentence, and a candidate in the form of a causal phrase from the sentence on which the pre-processing is performed extracting a triple, measuring a similarity using a causality detection model preset for the extracted candidate triple, and extending the candidate triple to a candidate quad including the measured similarity, and the measured similarity is preset If it is greater than or equal to the threshold, generating a patient health summary through semantic analysis using preset medical information, wherein the candidate triple is configured in the form of <noun, verb, noun>, and the candidate quad is <noun, It is composed of verb, noun, similarity> form.

According to the present invention, a medical service provider can easily grasp a patient's health condition.

In addition, the time it takes for the health care provider to process the patient's clinical record is shortened, thereby saving the health care provider's time.

It can also provide patients with full health status, increasing health care equity and efficiency.

In addition, it is possible to increase patient care ability by supporting clinical decision-making based on various information.

1 is a schematic structural block diagram of an apparatus for generating a patient health summary according to an embodiment of the present invention.
FIG. 2 is a detailed block diagram of the preprocessing unit shown in FIG. 1 .
FIG. 3 is a detailed block diagram of the main processing unit shown in FIG. 1 .
4 is a diagram illustrating an example of a configuration form of a health summary generated according to an embodiment of the present invention.
5 is a detailed block diagram of the candidate triple extractor shown in FIG.
6 is a detailed block diagram of the causal candidate classifier shown in FIG. 3 .
FIG. 7 is a detailed configuration block diagram of the semantic analyzer shown in FIG. 3 .
8 is a schematic flowchart of a method for generating a patient health summary according to an embodiment of the present invention;
9 is a diagram illustrating a schematic configuration of an apparatus for generating a patient health summary according to another embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, the embodiments of the present invention will be described in detail so that those of ordinary skill in the art to which the present invention pertains can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part "includes" a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated. In addition, terms such as “…unit”, “…group”, and “module” described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software or a combination of hardware and software. there is.

The devices described in the present invention are composed of hardware including at least one processor, a memory device, a communication device, and the like, and a program to be executed in combination with the hardware is stored in a designated place. The hardware has the configuration and capability to implement the method of the present invention. The program includes instructions for implementing the method of operation of the present invention described with reference to the drawings, and is combined with hardware such as a processor and a memory device to execute the present invention.

1 is a schematic structural block diagram of an apparatus for generating a patient health summary according to an embodiment of the present invention.

Referring to FIG. 1 , an apparatus 100 for generating a patient health summary according to an embodiment of the present invention includes a storage unit 110 , an input unit 120 , a preprocessor 130 , and a main processing unit 140 .

The storage unit 110 stores clinical records for each of one or more patients. Such clinical records may be recorded as data in a predetermined format. In addition, the amount of clinical records may vary according to the number of hospital examinations or conditions of the patient.

Also, the storage unit 110 may store the patient health summary output from the main processing unit 140 .

The input unit 120 receives the clinical record of a target patient for which a patient health summary is to be generated and provides it to the preprocessor 130 .

The input unit 120 may receive the patient's clinical record stored in the storage unit 120, but may also receive the patient's clinical record from the outside.

When a clinical record of a patient is input from the outside, the input unit 120 may store the inputted clinical record in the storage unit 110 .

The pre-processing unit 130 performs a pre-processing operation of converting the raw clinical text included in the patient's clinical record input from the input unit 120 into a format capable of subsequent processing. For example, the preprocessing operation performed by the preprocessor 130 may include a sentence extraction operation, a normalization operation, a tokenization operation, and the like. The configuration of the preprocessor 130 will be described in detail later.

The main processing unit 140 extracts a candidate phrase in the form of a causal phrase from the clinical record in which the pre-processing operation is performed by the pre-processing unit 130, and then generates a patient health summary through semantic analysis using existing medical information.

The patient health summary generated by the main processing unit 140 is output to the outside through the output unit 150 .

Also, the main processing unit 140 may store the generated patient health summary in the storage unit 110 .

The output unit 150 may output the patient health summary generated by the main processing unit 140 to the outside in a preset form or transmit it to an external device.

FIG. 2 is a detailed block diagram of the preprocessing unit 130 shown in FIG. 1 .

As shown in FIG. 2 , the preprocessor 130 includes a sentence extractor 131 , a regularizer 132 , a tokenizer 133 , a lemma extractor 134 , a stopword filter 135 and a tagging unit 136 . include

The sentence extractor 131 reads the clinical text included in the clinical record input from the input unit 120 and extracts it as a sentence. The clinical record input from the input unit 120 includes complete information about the patient expressed as a combination of words having a collective meaning. Accordingly, the sentence extractor 131 divides and extracts the word combination of the clinical record into small pieces of information called sentences.

For the sentences extracted by the sentence extractor 131, the normalizer 132 includes duplicate text, irrelevant information (eg, explanations in parentheses useful for the patient but not necessary for setting the situation, etc.), special characters (- , +, etc.) are removed. Since the sentence extracted by the sentence extractor 131 may include irrelevant information, the normalizer 132 performs a process to consist only of the relevant information.

The tokenizer 133 performs tokenization processing of dividing a word collection into tokens on the sentences normalized by the normalizer 132 . Since such tokenization processing is already well known, a detailed description thereof will be omitted here.

The lemma extractor 134 converts each token tokenized by the tokenizer 133 into a basic form. In general, in natural language, a word can have multiple representations, which increases the complexity of term processing. However, since the basic meaning of the terms remains the same, lemmatization is already well known to reduce complexity, so a detailed description will be omitted here.

The stop word filter 135 filters insignificant words, ie, stop words, which play a secondary role in the sentence extracted by the headword extractor 134 .

The tagging unit 136 uses a natural language processing parser, which is a syntax analyzer that processes natural language for sentences whose stopwords are filtered by the stopword filter 135, to attach parts of speech to each token (Part of Speech, POS) tagging is performed.

As such, the content processed by the aforementioned preprocessor 130 is a semantically enhanced token of a sentence that can be further processed by other components to generate a patient health summary.

Meanwhile, the preprocessor 130 may further include a converter (not shown) that converts the sentences on which the lemma extraction has been performed by the lemma extractor 134 into a specific character form. For example, when the clinical texts described in the clinical record are written in English, a converter for integrating all tokens included in the sentences output by the lemma extractor 134 into lowercase letters may be further included.

FIG. 3 is a detailed block diagram of the main processing unit 140 shown in FIG. 1 .

3 , the main processing unit 140 includes a candidate triple extractor 141 , a causal candidate classifier 142 , a semantic analyzer 143 , and a summary generator 144 .

Some terms used in the clinical record may be more important than others in relation to patient health. Accordingly, patient health related terms need to be distinguished from other terms, and the candidate triple extractor 141 is used to find all possible candidate terms within the sentences of the clinical record. The most important information in the clinical record is provided as a noun phrase followed by a combination of verb phrases and noun phrases.

Accordingly, the candidate triple extractor 141 extracts a candidate triple in the form of a <noun, verb, noun> phrase existing in sentences of the clinical record pre-processed and output by the pre-processing unit 130 .

The causality candidate classifier 142 measures the similarity of the candidate triple extracted by the candidate triple extractor 141 using the preset causality detection model 145, and then the measured similarity is greater than or equal to the preset threshold. Classify as a causality candidate, which is a quad containing only triples and similarities. For example, the quad may include the form of <noun, verb, noun, similarity>.

On the other hand, the above-described causality detection model 145 may be stored in the storage unit 110, a well-known artificial intelligence technology, for example, a deep learning technology in a triple consisting of <noun, verb, noun> Triples can be formed through learning where each noun represents a 'cause' and a 'effect'.

Since some of the quads that are causal candidates classified by the causal candidate classifier 142 may not be related to the clinical domain and may represent general information, quads that are not related to the clinical domain should be filtered out.

The semantic analyzer 143 filters the non-clinical quads that are not related to the clinical domain by determining the semantic type of each term constituting the quad using the existing medical dictionary. That is, the semantic analyzer 143 semantically enhances the information provided by the identified quad for the patient health summary. This semantic analysis shows the domain information of each quad.

The summary generator 144 groups the clinical quads semantically enhanced by the semantic analyzer 143 to generate a patient health summary and provides them to the output unit 150 and the storage unit 110 . At this time, the summary generator 144 may include all quad information in the patient health summary, but in order to provide a more summarized and simplified form of the patient health summary, for example, as shown in FIG. 4 , the quad A patient health summary including a summary of clinical information is created in the form of nouns representing 'cause' and 'effect' and 'reliability' corresponding to similarity among information. In other words, the patient health summary can be composed as the 'effect' corresponding to the 'cause' and reliability information about it.

As described above, according to the apparatus for generating a patient health summary according to an embodiment of the present invention, by providing a clinical record including clinical information for each patient as a corresponding patient health summary, a doctor who treats a patient, etc. can view the patient's health status at a glance As it helps to understand, it can save time for healthcare professionals, including doctors, and help understand patients.

5 is a detailed block diagram of the candidate triple extractor 141 shown in FIG.

As shown in FIG. 5 , the candidate triple extractor 141 includes a candidate triple identifier 1411 and a vector transformer 1412 .

The candidate triple identifier 1411 identifies a triple existing in sentences of the clinical record pre-processed and output by the pre-processing unit 130, that is, the form of a <noun, verb, noun> phrase as a candidate triple.

On the other hand, the candidate triple identified by the candidate triple identifier 1411 cannot be directly processed by the machine, and an encoding that can be understood by the machine is required.

The vector converter 1412 encodes the triples identified in the candidate triple identifier 1411 in vector form using word embedding techniques. Here, the word embedding technique is a method of converting a word into a dense representation, and is well known as a technique mainly used in natural language processing, so a detailed description thereof will be omitted here.

FIG. 6 is a detailed block diagram of the causal candidate classifier 142 shown in FIG. 3 .

As shown in FIG. 6 , the causal candidate classifier 142 includes a similarity identifier 1421 and a quad filter 1422 .

The similarity identifier 1421 measures the similarity using the causal detection model 145 for each of the candidate triples extracted by the candidate triple extractor 141 . Using this similarity measure, the proximity and relevance of a candidate to an already known term can be identified. In an embodiment of the present invention, the similarity identifier 1421 measures the similarity for each of the candidate triples using the cosine similarity, but is not limited thereto. Here, the cosine similarity is one of the methods of obtaining the similarity of text data, and is a method of obtaining the mutual similarity by obtaining the cosine angle between two vectors. Since the cosine similarity determines the similarity only by the angle regardless of the size of the data, it is possible to determine the similarity of the data even when the data set length difference is severe.

Also, the similarity identifier 1421 adds the measured maximum similarity to the candidate triple and expands it to a quad in the form of <noun, verb, noun, similarity>. Here, the maximum similarity refers to a similarity having a maximum value among the calculated similarities because a similarity with triples included in the causality detection model 145 is calculated for a candidate triple.

The quad filter 1422 classifies a quad including a triple and similarity as a causal candidate only for candidate triples in which the similarity measured for each candidate triple by the similarity identifier 1421 is equal to or greater than a preset threshold. Here, the higher the threshold value, the higher the accuracy, and the number of quads included in the summary may be limited. Therefore, the threshold may be set to an appropriate value through a statistical, experimental, or other method.

7 is a detailed configuration block diagram of the semantic analyzer 143 shown in FIG.

As shown in FIG. 7 , the semantic analyzer 143 includes a semantic category identifier 1431 and a medical quad filter 1432 .

The semantic category identifier 1431 determines the semantic type of each term using the existing medical dictionary for the candidate quads identified in the causal candidate classifier 142 . Here, a Unified Medical Language System (UMLS) may be used as an existing medical dictionary, but is not limited thereto. Here, UMLS is an integrated concept system that links concepts included in more than 60 medical thesaurus and taxonomy, and uses concepts rather than words and terms as basic entries. It is a source of knowledge to promote.

The medical quad filter 1432 performs filtering to remove non-clinical quads, not clinical quads, so that information of candidate triples whose semantic type is determined by the semantic category identifier 1431 is limited to a clinical profile. That is, the medical quad filter 1432 outputs only clinical quads related to the clinical domain.

Hereinafter, a method for generating a patient health summary according to an embodiment of the present invention will be described with reference to the description of the apparatus 100 for generating a patient health summary.

8 is a schematic flowchart of a method for generating a patient health summary according to an embodiment of the present invention;

Before the explanation, for example, using the terms used in the clinical profile using artificial intelligence technology, learn a triple in which two nouns represent 'cause' and 'effect', respectively, from a triple composed of <noun, verb, noun> It is assumed that the causality detection model 145 is determined in advance. Also, the patient health summary method according to an embodiment of the present invention may be performed by the patient health summary generating apparatus 100 described with reference to FIGS. 1 to 7 .

Referring to FIG. 8 , first, a clinical record corresponding to a specific patient is input ( S100 ). As described above, the clinical record for such a patient may be stored in the storage unit 110 or input from the outside.

Next, the clinical text included in the input clinical record is read to extract the corresponding sentences (S110), and it is determined whether the extracted sentence exists (S120).

If there is no sentence to be extracted, the patient health summary generating method according to the embodiment of the present invention is terminated. However, if at least one sentence exists, pre-processing is performed on the corresponding sentence (S130). This preprocessing involves normalization for the extracted sentences to remove phrases such as duplicate text, extraneous information (e.g., comments in parentheses useful to the patient but not necessary for setting the situation), and special characters (-, +, etc.) Processing, tokenization processing that divides the word collection into tokens for the normalized sentences, lemma extraction processing that converts each tokenized token into a basic form, insignificant role that plays a secondary role in the lemma extracted sentence Stopword filtering processing to filter out words, that is, stopwords, Part of Speech (POS) tagging processing that attaches parts of speech to each token using a natural language processing parser, a syntax analyzer that processes natural language for sentences filtered by stopwords, etc. is included, and since it has already been described above, a detailed description thereof will be omitted.

Next, a triple in the form of a <noun, verb, noun> phrase existing in the preprocessed sentences is identified as a candidate triple (S140).

Thereafter, it is determined whether the identified candidate triple exists ( S150 ). If the candidate triple does not exist, the patient health summary generating method according to the embodiment of the present invention is terminated.

However, if it is determined that at least one candidate triple exists, encoding is first performed in a vector form using a word embedding technique on one candidate triple ( S160 ).

Subsequently, similarity is measured using the causality detection model 145 for the vector-encoded candidate triple ( S170 ). In this case, the measured similarity may be measured in plurality for one candidate triple.

Thereafter, the candidate triples are expanded to quads to include the similarity having the maximum value among the measured similarities ( S180 ). Here, the quad may be expanded to a form in which the maximum degree of similarity is added to the triple, that is, <noun, verb, noun, similarity>.

Next, it is determined whether the maximum similarity among the similarities measured in step S170 is equal to or greater than a preset threshold value (S190).

If the similarity, that is, the maximum similarity is greater than or equal to a preset threshold, the semantic category is identified using a medical dictionary such as UMLS (Unified Medical Language System) for the quad extended to the causality candidate (S200), and the identified semantic category is It is determined whether it is a clinical quad related to a clinical domain (S210).

If it is determined that the clinical quad is related to the clinical domain through the identified semantic category, the corresponding candidate quad is added to the patient health summary ( S220 ). At this time, the quad information in the form of <noun, verb, noun, similarity> is added to the patient health summary in the form of <cause -> result, reliability> as shown in FIG. 4 . Here, since 'cause' and 'result' have already been determined in the above-described step (S170), whether or not causality is determined is replaced with a noun corresponding to 'cause' and a noun corresponding to 'result' among the two nouns, and 'reliability' is replaced by the similarity of the quad.

In this way, when the addition to the patient health summary corresponding to one candidate triple identified in step 140 is completed, the addition to the patient health summary can be performed for other candidate triples through the same process. Steps S150 to S220 are repeatedly performed.

Meanwhile, in the above-described step (S190), when the similarity, that is, the maximum similarity, is smaller than a preset threshold, and in the above-described step (S210), even if it is not a clinical quad, the above steps (S150 to S220) for other candidate triples. The procedure is in progress so that it can be carried out.

In this way, when the above-described steps (S150 to S220) are performed for all candidate triples, only the patient health summary corresponding to the clinical quad is added to the patient health summary, so the final patient health summary is stored in the storage unit 110 and It may also be provided to a doctor or the like.

9 is a diagram illustrating a schematic configuration of an apparatus for generating a patient health summary according to another embodiment of the present invention.

As shown in FIG. 9 , the apparatus 200 for generating a patient health summary according to another embodiment of the present invention includes at least one processor 210 , a memory 220 , a communicator 230 , an input/output device 240 , and a communication device. Bus 250 is included.

The processor 210 may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling program execution in the solution of the present application.

The memory 220 stores information related to generating a patient health summary according to an embodiment of the present invention.

Specifically, the memory 220 is further configured to store a set of codes, and the codes are used to control the processor 210 to execute a process as follows. These processes include a process of extracting at least one sentence from the contents of clinical information included in a clinical record for a patient, a process of performing preprocessing for performing part-of-speech tagging on the extracted sentence, and a causal phrase from the preprocessed sentence. A process of extracting a candidate triple of the shape, a process of measuring the similarity using a causality detection model preset for the extracted candidate triple, a process of expanding the candidate triple into a candidate quad including the measured similarity, and the measured similarity is and a process of generating a patient health summary through semantic analysis using the preset medical information when it is above a preset threshold.

Here, the candidate triple is configured in the form of <noun, verb, noun>, and the candidate quad is configured in the form of <noun, verb, noun, similarity>.

In addition, the patient health summary is configured in the form of <cause -> result, reliability>, cause and effect constitute a candidate quad, and among the two nouns indicating causality, the noun indicating the cause is the cause, and the noun indicating the effect is set as the result, and the reliability is replaced with the degree of similarity constituting the candidate quad.

In addition, this process further includes a process of receiving a clinical record from the outside through the input/output unit 240 , and a process of outputting the generated patient health summary to the outside.

In addition, the process further includes a process of receiving a clinical record from the outside through the communicator 230 , and a process of outputting the generated patient health summary to the outside.

In addition, the process of extracting a candidate triple in the form of a causal phrase from the sentence on which the preprocessing has been performed includes a process of performing vector encoding on the candidate triple by using a word embedding technique.

Memory 220 can be read-only memory (ROM) or other type of static storage that can store instructions, or random access memory (RAM) or other type of dynamic storage that can store information and instructions; or Electrically Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM), or other compact disc storage device or optical disc storage device (compressed optical disc, laser disc, optical disc, digital versatile disc, blue ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can carry or store expected program code in the form of instructions or data structures and that can be accessed by a computer; This is not limited. The memory 220 may exist independently and is coupled to the processor 210 by a communication bus 250 .

The communicator 230 communicates with other devices or communication networks, and may be implemented using various communication technologies. That is, Wi-Fi (WIFI), WCDMA (Wideband CDMA), HSDPA (High Speed Downlink Packet Access), HSUPA (High Speed Uplink Packet Access), HSPA (High Speed Packet Access), Mobile WiMAX (Mobile WiMAX), WiBro (WiBro) , LTE (Long Term Evolution), Bluetooth (bluetooth), infrared data association (IrDA), NFC (Near Field Communication), Zigbee, wireless LAN technology, USB (Universal Serial Bus), etc. can be applied. . In addition, when a service is provided by being connected to the Internet, TCP/IP, which is a standard protocol for information transmission on the Internet, may be followed.

The input/output device 240 specifically includes an input device 241 and an output device 242 , and the input device 241 may communicate with the processor 210 and receive user input in a plurality of ways. . For example, the input device 241 may be a mouse, a keyboard, a touch screen, or a sensing device. The output device 242 may communicate with the processor 210 and display information or output voice in a plurality of ways. For example, the output device 242 may be a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, a speaker, or the like.

The communication bus 250 is configured to couple all the components of the apparatus 200 for generating a patient health summary, namely the processor 210 , the memory 220 , the communicator 230 and the input/output 240 .

The embodiment of the present invention described above is not implemented only through the apparatus and method, and may be implemented through a program for realizing a function corresponding to the configuration of the embodiment of the present invention or a recording medium in which the program is recorded.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto. is within the scope of the right.

Claims (13)

A device for generating a patient health summary, comprising:
A preprocessor that extracts at least one sentence from the contents of clinical information included in the clinical record for a patient, and performs preprocessing for performing part-of-speech tagging on the extracted sentence, and
A main processing unit that extracts a candidate phrase in the form of a causal phrase from a sentence that has been pre-processed by the pre-processing unit, and then generates a patient health summary through semantic analysis using preset medical information.
includes,
The candidate syntax of the causal syntax form represents a triple of the form <noun, verb, noun> identified from the sentence on which preprocessing has been performed by the preprocessor,
Patient health summary generation device. According to claim 1,
The main processing unit,
a candidate triple extractor for identifying triples existing in the sentence on which the preprocessing has been performed by the preprocessor and extracting them as candidate triples;
When the candidate triple extracted by the candidate triple extractor is a causal triple in which two nouns constituting the candidate triple form a causal relationship representing a cause and an effect, respectively, a causal candidate classifier for classifying the candidate triple as a causal candidate ,
A semantic analyzer for filtering causal candidates corresponding to clinical domains by determining semantic types of terms constituting the causal candidates classified by the causal candidate classifier; and
Summary generator that groups the causal candidates filtered by the semantic analyzer to generate the patient health summary
A device for generating a patient health summary comprising a. 3. The method of claim 2,
The candidate triple extractor is
A candidate triple identifier for identifying triples present in the sentence on which the preprocessing has been performed by the preprocessor, and
A vector converter that performs vector encoding on a candidate triple identified by the candidate triple identifier using a word embedding technique
A device for generating a patient health summary comprising: 3. The method of claim 2,
The causal candidate classifier is
A similarity identifier that measures the similarity using a causal detection model preset with respect to the candidate triple extracted by the candidate triple extractor, and expands the candidate triple into a candidate quad including the measured similarity, and
A quad filter for filtering only candidate quads whose similarity measured by the similarity identifier is equal to or greater than a preset threshold value as the causal candidates
includes,
The candidate quad is configured in the form of <noun, verb, noun, similarity>,
Patient health summary generation device. 5. The method of claim 4,
The causality detection model uses artificial intelligence technology to learn to include a causal triple in which two nouns represent a causal relationship corresponding to a 'cause' and a 'effect' in a triple composed of a <noun, verb, noun> form, respectively. formed,
Patient health summary generation device. 5. The method of claim 4,
The similarity identifier uses a cosine similarity method when measuring the similarity for the candidate triple using the causality detection model.
Patient health summary generation device. 5. The method of claim 4,
The similarity included in the candidate quad is a degree of similarity having a maximum value among one or more similarities measured by the similarity identifier with respect to the candidate triple,
Patient health summary generation device. 3. The method of claim 2,
The semantic analyzer is
a semantic category identifier for identifying a semantic type of a term using a medical dictionary preset for a causal candidate classified by the causal candidate classifier; and
Medical quad filter for removing non-clinical candidates whose semantic type identified by the semantic category identifier does not correspond to the clinical domain
A device for generating a patient health summary comprising: 5. The method of claim 4,
The summary generator generates a patient health summary in the form of <cause -> result, reliability> from the form of <noun, verb, noun, similarity> constituting the candidate quad filtered by the semantic analyzer,
The cause and effect constitute the candidate quad, and a noun indicating a cause among two nouns indicating a causal relationship is the cause, and a noun indicating a result is set as the result, and the reliability is the degree of similarity constituting the candidate quad. replaced by
Patient health summary generation device. According to claim 1,
The preprocessor is
a sentence extractor for extracting sentences from the contents of clinical information included in the clinical record;
A normalizer that removes duplicate text and content containing special characters for the sentences extracted by the sentence extractor;
A tokenizer that performs tokenization processing of dividing a collection of words into tokens for the sentences normalized by the normalizer;
A lemma extractor that converts each token tokenized by the tokenizer into a basic form;
A stopword filter for removing stopwords from the sentences converted by the headword extractor, and
A tagging unit for tagging each token with a part-of-speech by using a natural language processing parser for the sentences filtered by the stopword filter
A device for generating a patient health summary comprising: A method of generating a patient health summary comprising:
extracting sentences from the contents of clinical information included in the clinical record for the patient;
performing pre-processing for performing part-of-speech tagging on the extracted sentence;
extracting a candidate triple in the form of a causal syntax from the sentence on which the preprocessing has been performed;
Measuring similarity using a causality detection model preset for the extracted candidate triple, and expanding the candidate triple to a candidate quad including the measured similarity; and
generating a patient health summary through semantic analysis using preset medical information when the measured similarity is greater than or equal to a preset threshold
includes,
The candidate triple is configured in the form of <noun, verb, noun>, and the candidate quad is configured in the form of <noun, verb, noun, similarity>,
How to create a patient health summary. 12. The method of claim 11,
The step of extracting a candidate triple in the form of a causal syntax from the sentence on which the pre-processing has been performed,
performing vector encoding on the candidate triple using a word embedding technique;
How to create a patient health summary. 12. The method of claim 11,
The patient health summary is configured in the form of <cause -> result, reliability>, wherein the cause and effect constitute the candidate quad, and the noun indicating the cause is the cause among the two nouns indicating the causal relationship, and the noun indicating the result is the cause. set a noun to the result, wherein the confidence is replaced by the degree of similarity constituting the candidate quad,
How to create a patient health summary.

Images