CN109657245B - A Semantic Recognition Method for Electronic Medical Records - Google Patents

Abstract

The present invention provides a semantic recognition method for electronic medical records, comprising the following steps: S1. Collect several electronic medical records to establish a professional corpus and a professional field dictionary, wherein the professional field dictionary includes several professional words, stop words and professional words in The word frequency in the professional corpus, the professional words include critical words and common words; establish a transfer form, the condition of the transfer form is at least one critical word, and the result is the target department; S2, extract the inspection content in the electronic medical record; S3. Perform word segmentation on the inspection content according to the professional domain dictionary; this electronic medical record semantic recognition method can accurately identify critical words in the electronic medical record that are closely related to related diseases, and then quickly find the corresponding words according to the transfer form. The target department has greatly improved the accuracy and efficiency of finding the target department.

Description

A Semantic Recognition Method for Electronic Medical Records

technical field

The invention relates to data processing of electronic medical records, in particular to a semantic recognition method of electronic medical records.

Background technique

With the development of information technology in the era, the use of electronic medical records has gradually become popular in medical activities. By analyzing and mining electronic medical records, a large amount of medical data about individual patient information can be obtained, and the critical value of these data is the part that most needs the attention of doctors and patients. The critical value refers to the abnormal result of a certain item or type of test. When the abnormal result of this test appears, it indicates that the patient may be in a dangerous state. Clinicians need to obtain test information in a timely manner and quickly give patients effective intervention measures and treatment. At present, the critical value is manually checked by a special medical technology department and the conclusion is reported to the corresponding department, and the doctor in the department makes a further diagnosis. This kind of diagnosis method has too high professional requirements for the inspectors. They are very proficient in the diseases of various subjects and cannot give a definite answer to the diagnosis results. The judgment they made when choosing a department may not be accurate. This will not only increase the communication cost and efficiency of the hospital, but more seriously may cause delay in treatment.

Contents of the invention

The technical problem to be solved by the present invention is: in order to solve the problem that the current manual judgment of electronic medical records not only consumes a lot of time and energy of doctors, but also the accuracy of judgment results cannot be guaranteed, the present invention provides a semantic recognition method for electronic medical records to solve above question.

The technical scheme that the present invention adopts to solve its technical problem is: a kind of semantic recognition method of electronic medical record, comprises the following steps:

S1. Collect several electronic medical records to establish a professional corpus and a professional field dictionary. The professional field dictionary includes several professional words, stop words and the frequency of professional words in the professional corpus. The professional words include critical words and common word; set up a transfer form, the condition of the transfer form is at least one critical word, and the result is the target department;

S2. Extract the examination content in the electronic medical record;

S3. Perform word segmentation on the inspection content according to the professional field dictionary;

S4. Establish evaluation function

其中

为平衡因子，MI(W)表示字串W中字与字之间的互信息，H_L(W)表示字串W的左信息熵，H_R(W)表示字串W的左信息熵；

in

For balance factor, MI (W) represents the mutual information between word and word in word string _W , HL (W) represents the left information entropy of word string W, _HR (W) represents the left information entropy of word string W;

S5, rank (W) is compared with threshold value P, if rank (W) ≥ P, then determine word string W is a word, if rank (W) < P, then determine word string W is not a word;

S6. Put all word strings W as conditions into the transfer form to search for the target department.

As preferably, also include step S7, inversely find out the critical words contained in the corresponding conditions through the target department found in the transfer form, negatively detect these critical words, if the result of the negative detection is that no negative is detected, the electronic The medical record is sent to the target department, and if the result is negative, the electronic medical record is marked as pending.

As preferably, the professional field dictionary also includes negative terms, and the semantic recognition method of the electronic medical record also includes the steps:

Set up a negative detection form, the conditions of the negative detection form include a negative term and at least one critical word, and the result is pending;

Separated by commas, extract the critical phrases that will contain critical words in the inspection report, and put all the words in the critical phrases into the negative detection form search results as conditions. If the result is pending, it means that the critical phrase is a negative description , the electronic medical record is marked as pending; if no result is found, it means that the critical phrase is a positive description, and if all the critical phrases are positive descriptions, the electronic medical record is sent to the target department.

Preferably, in step S5, if it is judged that the word string W is a word, then add the word string W as a new professional word into the professional field dictionary.

The beneficial effect of the present invention is that the semantic recognition method for electronic medical records can accurately identify critical words closely related to related diseases in electronic medical records, and then quickly find the corresponding target department according to the transfer form, which greatly improves the efficiency of finding the target department. accuracy and efficiency.

Description of drawings

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

Fig. 1 is a flowchart of an embodiment of a semantic recognition method for electronic medical records.

Fig. 2 is a flow chart of another embodiment of a semantic recognition method for electronic medical records.

Detailed ways

Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

In describing the present invention, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", " Back", "Left", "Right", "Vertical", "Horizontal", "Top", "Bottom", "Inner", "Outer", "Axial", "Radial", "Circumferential", etc. The indicated orientation or positional relationship is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, or in a specific orientation. construction and operation, therefore, should not be construed as limiting the invention.

In addition, the terms "first", "second", etc. are used for descriptive purposes only, and should not be construed as indicating or implying relative importance. In the description of the present invention, it should be noted that unless otherwise specified and limited, the terms "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral Ground connection; it can be mechanical connection or electrical connection; it can be direct connection or indirect connection through an intermediary. Those of ordinary skill in the art can understand the specific meanings of the above terms in the present invention in specific situations. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.

Any process or method descriptions in flowcharts or otherwise described herein may be understood to represent modules, segments or portions of code comprising one or more executable instructions for implementing specific logical functions or steps of the process , and the scope of preferred embodiments of the invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including substantially concurrently or in reverse order depending on the functions involved, which shall It is understood by those skilled in the art to which the embodiments of the present invention pertain.

As shown in Figure 1, it is an embodiment of a semantic recognition method of an electronic medical record, comprising the following steps:

S1. Collect several electronic medical records to establish a professional corpus and a professional field dictionary. The professional field dictionary includes several professional words, stop words and the frequency of professional words in the professional corpus. The professional words include critical words and common Words; critical words such as mobile dullness, fatigue, anorexia, yellow urine, oozing blood, cough, hemoptysis, etc. Common words are other words commonly used in medical records, and stop words are words that have no role in judging symptoms in electronic medical records , such as bar, ah, unless, and etc.; set up a transfer form, the condition of the transfer form is at least one critical word, and the result is the target department; Table 1 provides an example of the transfer form:

critical word target department coughing, coughing up blood Respiratory hydroarthrosis Orthopedics stone internal medicine mouth ulcers Dentistry

Table 1S2, extracting the inspection content in the electronic medical record;

S3, use the jieba word segmentation tool to perform word segmentation on the inspection content according to the professional field dictionary, remove stop words after word segmentation to obtain several strings W, and establish a formula:

Among them, MI(W) represents the mutual information between all words in the word string W, and the word string W is composed of n words in total, and x ₁ , x ₂ , x ₃ ,…, x _n represent the first 1 word, the 2nd word, the 3rd word...the nth word, N(x _n ) indicates the frequency of occurrence of the word x _n in the professional corpus, N(x ₁ ,x ₂ ,x ₃ ,…,x _n ) Indicates the frequency with which the word string W appears in the professional corpus as a whole;

S4. Establish evaluation function

其中

为平衡因子，MI(W)表示字串W中字与字之间的互信息，H_L(W)表示字串W的左信息熵，H_R(W)表示字串W的左信息熵；

in

For balance factor, MI (W) represents the mutual information between word and word in word string _W , HL (W) represents the left information entropy of word string W, _HR (W) represents the left information entropy of word string W;

Among them, p( _xi W) and p( _Wxi ) represent the conditional probability when the word x _i appears on the left and right of W in the professional corpus, _xi represents the professional corpus, V _L and V _R represent A set of all words appearing on the left and right of W in the professional corpus;

S5, rank (W) is compared with the preset threshold value P, if rank (W) ≥ P, then it is determined that the word string W is a word, if rank (W) < P, then it is determined that the word string W is not a word and Abandon it; if it is judged that word string W is a word, then this word string W is added in described specialized field dictionary as new specialized word;

S6. Put all word strings W as conditions into the transfer form to search for the target department.

Considering that negative descriptions often appear in the phrases where critical words are located in medical records, such critical situations do not appear in the expression, so in some other examples, as shown in Figure 2, steps are also included:

S7. The critical words contained in the corresponding conditions are reversely checked by the found target department in the transfer form, and these critical words are negatively detected: a negative detection form is established, the professional field dictionary includes negative terms, and the conditions of the negative detection form Including a negative term and at least one critical word, the result is pending. For example, Table 2 shows part of a negative detection form:

negative term critical word result none Fatigue, poor appetite, yellow urine to be determined none Lip cyanosis, bleeding gums to be determined did not appear coughing, coughing up blood to be determined unheard of dry rales, wet rales to be determined deny tumor, nodule to be determined not accompanied joint swelling to be determined exclude SARS to be determined feminine shifting dullness to be determined

Table 2

Separated by commas, extract the critical phrases that will contain critical words in the inspection report, and put all the words in the critical phrases into the negative detection form search results as conditions. If the result is pending, it means that the critical phrase is a negative description , the electronic medical record is marked as pending; if it is the result of the query, it means that the critical phrase is a positive description, and if all the critical phrases are positive descriptions, the electronic medical record is sent to the target department.

In the description of this specification, descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific examples", or "some examples" mean that specific features described in connection with the embodiment or example , structure, material or feature is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Inspired by the above-mentioned ideal embodiment according to the present invention, through the above-mentioned description content, relevant workers can make various changes and modifications within the scope of not departing from the technical idea of the present invention. The technical scope of the present invention is not limited to the content in the specification, but must be determined according to the scope of the claims.

Claims (4)

1. A semantic recognition method of an electronic medical record is characterized by comprising the following steps:

s1, collecting a plurality of electronic medical records to establish a professional corpus and a professional field dictionary, wherein the professional field dictionary comprises a plurality of professional words, stop words and word frequency of the professional words in the professional corpus, and the professional words comprise critical words and common words; establishing a transfer form, wherein the condition of the transfer form is at least one critical word, and the result is a target department;

s2, extracting the inspection content in the electronic medical record;

s3, segmenting the checked content according to the professional field dictionary;

s4, establishing an evaluation function

Wherein

MI (W) represents the word-to-word mutual information in the string W, H, as a balance factor _L (W) left entropy, H, of word string W _R (W) left entropy representing string W;

s5, comparing rank (W) with a threshold value P, if rank (W) is more than or equal to P, determining that the word string W is a word, and if rank (W) < P, determining that the word string W is not a word;

and S6, putting all word strings W as conditions into a forwarding table to search a target department.

2. The method for recognizing the semantic meaning of the electronic medical record as claimed in claim 1, wherein: and S7, reversely checking the searched target department in the forwarding form to obtain critical words contained in the corresponding conditions, carrying out negative detection on the critical words, sending the electronic medical record to the target department if the negative detection result is that the negative detection is not carried out, and marking the electronic medical record to be processed if the negative detection result is that the negative detection is not carried out.

3. The method for recognizing the semantic meaning of the electronic medical record as claimed in claim 2, wherein: the professional domain dictionary further comprises negative terms, and the method for recognizing the semanteme of the electronic medical record further comprises the following steps:

establishing a negative detection form, wherein the conditions of the negative detection form comprise a negative term and at least one critical word, and the result is pending;

taking commas as separation, extracting critical phrases containing critical words in the inspection report, putting all words obtained by word segmentation in the critical phrases as conditions into a negative detection form to search for results, and marking the critical phrases as negative description if the results are undetermined, so as to mark the electronic medical record as pending; if the result is not found, the emergency phrase is represented as a positive description, and if all the emergency phrases are positive descriptions, the electronic medical record is sent to the target department.

4. The method as claimed in claim 3, wherein the semantic recognition of the electronic medical record comprises: in step S5, if the word string W is judged to be a word, the word string W is added as a new professional word to the professional domain dictionary.

Images