CN112749307A - Medical data processing method and device and storage medium - Google Patents
Medical data processing method and device and storage medium Download PDFInfo
- Publication number
- CN112749307A CN112749307A CN202011604300.2A CN202011604300A CN112749307A CN 112749307 A CN112749307 A CN 112749307A CN 202011604300 A CN202011604300 A CN 202011604300A CN 112749307 A CN112749307 A CN 112749307A
- Authority
- CN
- China
- Prior art keywords
- information
- original
- standard
- medical data
- operation information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The present disclosure relates to a processing method of medical data, a processing apparatus of medical data, and a computer-readable storage medium, the processing method including acquiring original medical data, the original medical data containing original surgery information; matching the original operation information with standard words of a standard operation information table to obtain a matching relation; screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship; and carrying out normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data. Through the embodiments of the disclosure, the medical data containing the operation information with complex description can be accurately and efficiently processed, so that the operation information can be intelligently processed.
Description
Technical Field
The present disclosure relates to the field of medical data intelligent processing technology, and in particular, to a medical data processing method, a medical data processing apparatus, and a computer-readable storage medium.
Background
In the process of processing medical data containing operation information, the prior art only realizes data processing and information sorting based on short text operation names. In medical data related to operation information, for example, a complete operation record generally includes an operation name and an operation pass, when an electronic medical record is written, because an operation has a lot of details, and some differences in habits or specifications are added, the recording results of different recording modes for the same operation are very different, wherein the situations that the operation details are lacked in the operation name or individual operations are lacked are very common, and even the situation that the operation name is wrong can occur more seriously.
Disclosure of Invention
The present disclosure is intended to provide a medical data processing method, a medical data processing apparatus, and a computer-readable storage medium, which can accurately and efficiently process medical data containing operation information of a complex description, thereby intelligently processing the operation information.
According to one aspect of the present disclosure, there is provided a method for processing medical data, including:
acquiring original medical data, wherein the original medical data comprises original operation information;
matching the original operation information with standard words of a standard operation information table to obtain a matching relation;
screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship;
and carrying out normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data.
In some embodiments, the first and second light sources, wherein,
the obtaining of the matching relationship comprises: combining the original operation information and each standard word of a standard operation information table in pairs, and respectively calculating matching degrees;
based on the matching relationship, screening out the to-be-processed operation information respectively containing the original operation information and the standard words, and the method comprises the following steps: and under the condition that the matching degree meets a preset condition, taking the original operation information and the standard words of the standard operation information table as the operation information to be processed.
In some embodiments, the normalizing the surgical information to be processed by the classification algorithm includes:
inputting the operation information to be processed into a neural network classification model;
analyzing whether the original operation information in the operation information to be processed corresponds to the standard words or not;
according to the analysis result, normalization processing is carried out on the corresponding original operation information and the standard words;
outputting standard surgical information corresponding to the original surgical information for generating the target medical data.
In some embodiments, the neural network classification model is obtained based on a pre-training mode, and the pre-training mode includes:
acquiring original medical data containing original surgical information;
extracting operation name information and operation passing information in the original operation information;
combining the operation passing information with standard words in an operation name library, wherein the operation name library is constructed based on the operation name information and comprises an original operation name and a standard operation name of original operation information;
training a pre-training model by using the combined surgical procedure information and standard words in a surgical name library as training data;
and obtaining the neural network classification model based on the pre-training model.
In some embodiments, the operation name library is constructed in a manner including:
inputting the operation name information into a neural network model for coding an operation name so as to extract a standard operation name code from the operation name information;
and constructing an operation name library containing the original operation names and the coded standard operation names based on the operation name information and the standard operation name codes.
In some embodiments, wherein obtaining the target medical data comprises:
characterizing the original surgical information in a standard surgical information format;
outputting the target medical data to update the raw medical data.
In some embodiments, among others, further comprising:
and analyzing the content of the original operation information to obtain an operation name and a long text operation process.
According to one aspect of the present disclosure, there is provided a processing apparatus of medical data, comprising:
an acquisition unit configured for acquiring raw medical data, the raw medical data containing raw surgical information;
the matching module is configured to match the original operation information with standard words of a standard operation information table to obtain a matching relation;
the screening module is configured to screen out to-be-processed surgical information respectively comprising original surgical information and standard words based on the matching relationship;
the processing module is configured to perform normalization processing on the surgical information to be processed through a classification algorithm to obtain target medical data.
According to one aspect of the present disclosure, there is provided a computer-readable storage medium having stored thereon computer-executable instructions that, when executed by a processor, implement:
the method for processing medical data according to the above.
The medical data processing method, the medical data processing device and the computer readable storage medium of various embodiments of the present disclosure are realized by acquiring original medical data, wherein the original medical data contains original surgery information; matching the original operation information with standard words of a standard operation information table to obtain a matching relation; screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship; and performing normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data, so that the target medical data can be generated based on a standard operation information table to process and sort the original medical data based on the original operation information, particularly the original medical data containing the operation process description content of the long text, to obtain a unified normalization processing result. According to the processing method, the idea and the method for extracting the standard operation name from the long text of the operation process can be applied to the scene of supplementing and correcting the original operation name by using the content of the operation process, the recall rate of the data processing performance from the upper end to the upper end reaches 95%, various operation information is completely, accurately and efficiently expressed by outputting the standard information and a uniform coding mode, and therefore the accuracy and the efficiency of medical research and medical diagnosis and treatment are improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure, as claimed.
Drawings
In the drawings, which are not necessarily drawn to scale, like reference numerals may designate like components in different views. Like reference numerals with letter suffixes or like reference numerals with different letter suffixes may represent different instances of like components. The drawings illustrate various embodiments generally, by way of example and not by way of limitation, and together with the description and claims, serve to explain the disclosed embodiments.
Fig. 1 shows a flow chart of a method of processing medical data to which an embodiment of the present disclosure relates;
fig. 2 shows an architecture diagram of a medical data processing apparatus according to an embodiment of the present disclosure.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present disclosure more clear, the technical solutions of the embodiments of the present disclosure will be described below clearly and completely with reference to the accompanying drawings of the embodiments of the present disclosure. It is to be understood that the described embodiments are only a few embodiments of the present disclosure, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the disclosure without any inventive step, are within the scope of protection of the disclosure.
Unless otherwise defined, technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this disclosure belongs. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items.
To maintain the following description of the embodiments of the present disclosure clear and concise, a detailed description of known functions and known components have been omitted from the present disclosure.
The present disclosure relates to the processing of medical data for the processing of surgical information against a raw record. In the process of processing medical data containing operation information, the prior art only realizes data processing and information sorting based on short text operation names, for example, realizes coding based on short text operation names, and cannot perform corresponding data processing based on long text, for example, aiming at texts with more than hundreds of characters. In medical data related to operation information, for example, a complete operation record generally includes an operation name and an operation pass, when an electronic medical record is written, because the operation has a lot of details and some differences in habits or specifications, different recording modes, for example, different doctors in different hospitals have a great difference in recording results of the same operation, wherein the case that the operation details are lacked or individual operations are lacked in the operation name is very common, and the case that the operation name is wrong is more serious, even occurs. Based on the existing difference, a more complete operation record needs to be generated, so as to unify the standardized output of the operation information to solve some problems encountered in practical clinics, such as supplementing and correcting the operation name.
As one aspect, as shown in fig. 1, an embodiment of the present disclosure provides a method for processing medical data, including:
s101: acquiring original medical data, wherein the original medical data comprises original operation information;
s102: matching the original operation information with standard words of a standard operation information table to obtain a matching relation;
s103: screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship;
s104: and carrying out normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data.
One of the inventive concepts of the present disclosure is to process and sort information of original medical data based on a standard operation information table according to original operation information, particularly original medical data of operation procedure description contents including a long text, so as to obtain a unified normalization processing result to generate target medical data.
The original medical data in the embodiments of the present disclosure, which belongs to the data source, need not be particularly limited, and may be historical data or current real-time data. From the aspect of data format, the medical record text data, video data, audio data and the like can be used as long as the original operation information which can be contained in the medical record text data, the original operation information can be identified by identification means, such as text identification (for example, NLP identification, OCR identification and the like), some medical information such as operation names and operation courses which are described in the medical record text data, the medical information content which is identified by character splitting, word splitting and the like, and the like can be identified. Referring to the ICD9-CM-3 standard information table, the raw medical data of the embodiments of the present disclosure includes a variety of operation information from which operation name information can be extracted and information describing the operation procedure, especially for long text content information, which may include hundreds and thousands of words, and may be recorded in the raw medical data based on the form of data entry. In some embodiments, the raw medical data of the present disclosure may also be medical records, diagnostic books, surgical reports, which include a plurality of or a plurality of surgical names and surgical descriptions, and these information may or may not include codes characterizing the respective information. In various data information analysis scenarios, the original medical data in the embodiment of the present disclosure may be a medical text of the original medical data input by a user through an interactive interface and an input device, and may be used for interpretation of relevant medical information by way of annotation or analysis, such as manual or machine.
In each embodiment, in the implementation process of the present disclosure, the operation name information and the operation pass description information in the original medical operation information of the present embodiment may be extracted through a neural network model. In the implementation process, the specific neural network model is not particularly limited, and can be implemented by adopting a neural network model which meets the requirements and is matched with the architecture. According to the more preferable scheme, the extraction accuracy of various information can be further optimized through the adaptive neural network model on the basis of the pre-training model. For extracting medical entity content, entity extraction can be performed based on a text recognition mode, for example, a text recognition mode such as NLP (natural language processing), and clauses and classifications are performed on entities by combining medical concepts. More preferably, the entity can be analyzed by combining with a standard medical information table, such as various information tables of ICD, and extracted on the basis of combining with a proper medical rule analysis result.
In some embodiments, the obtaining a matching relationship of the present disclosure includes: combining the original operation information and each standard word of a standard operation information table in pairs, and respectively calculating matching degrees;
based on the matching relationship, screening out the to-be-processed operation information respectively containing the original operation information and the standard words, and the method comprises the following steps: and under the condition that the matching degree meets a preset condition, taking the original operation information and the standard words of the standard operation information table as the operation information to be processed.
Specifically, the standard surgical information table of the present embodiment can be obtained through the ICD-9-CM-3 standard table, and the surgical information can be calibrated in detail and the corresponding surgical code can be given through the ICD-9-CM-3 standard table.
The original operation information in the original medical data and the standard medical information in the ICD-9-CM-3 standard table, or called standard words, are combined in pairs. For example, the input in this embodiment is surgical case history data, wherein the original text records the surgical information as
"name of operation: coronary angiography + PCI
The operation process comprises the following steps: the operation area is disinfected conventionally, a sterile hole towel is laid, the right radial artery puncture is successful, and the radiography shows that: the left trunk is normal, the front descending branch is 7-8 sections and is 80% narrow, the circumgyrating branch is 11 sections and is 80% narrow, the right crown is 2-3 sections and is 40% narrow, and the PDA opening is 70% narrow. PCI (peripheral component interconnect) operation: replacing AL1guiding6F to the left coronary orifice, feeding a BMW guide wire to the distal end of the anterior descending branch, feeding another BMW to the distal end of D2, feeding a Tazuna2.0 × 15mm saccule to expand the 7-8 segment lesion at 12-13atm for 7-8 seconds, then implanting a Firenbird3.0 × 20mm scaffold to expand the stent at 14atm for 8 seconds, releasing the 3.5 × 18mm scaffold at 7 segment 12atm8 seconds, expanding the saccule expandable stent at 3.0 × 10mm for 18atm7-8 seconds, expanding the circumflex-branch lesion at 2.0 × 15mm saccule from the BMW guide wire to the distal end of the circumflex-branch for 16atm8 seconds, releasing the 3.5 × 15mm scaffold at 12atm8 seconds, repeatedly showing good stent expansion by radiography, having no residual stenosis, achieving the TIMI3 grade of blood flow, removing the catheter and the sheath, locally compressing hemostasis, finishing the surgery,
the name of the operation is only written as 'PCI operation', but according to the recorded content of the operation and the medical knowledge, the type of the stent can be determined, several stents are actually implanted, and the patient also has percutaneous coronary artery sacculus dilatation and plasty. Then, based on the data processing method of the present disclosure, at least error correction of the procedure name according to the procedure pass can be realized for the input.
The standard surgical information table of this embodiment includes a plurality of rows of data entries containing information such as primary code, additional code, surgical name, category, etc. And combining the input text original text and each standard word in the standard operation information table in pairs to form a data item in the format of 'original text and standard word'. E.g. forming
"1. (original, non-drug eluting coronary stents placed)
2. (original, coronary artery drug coating stent implantation)
3. (original, drug eluting coronary stent)
… …' (pair of information).
The operation passing information is described through text, so that the operation passing information is the detailed content of the operation name, and the operation passing information and the operation name have a corresponding relation. Based on the correspondence, each pair of combined pair can be used as a piece of training data in this embodiment.
The matching degree is calculated for each pair combination of the above forms, and the combination with the matching degree lower than a preset threshold is discarded and is not used as a data normalization target. And taking the combination with the matching degree meeting the preset threshold as a to-be-selected normalization object.
Specifically, the determination method of the matching degree in this embodiment may be calculated based on the number of words included in each of the original text and the standard word, for example, the matching degree is calculated by the number of words in intersection between the original text and the standard word and the number of words in union of the original text and the standard word, or is referred to as the similarity between the original text and the standard word. By way of example with simple text numbers, the original text contains 'bilateral thyroidectomy', a standard word after two-two combination is 'thyroidectomy', the intersection of the two is 'thyroidectomy', the number of words is 4, and the union of the two is the sum of the lengths of two character strings minus the number of intersection words: 9+8-4 is 13, so the similarity of this pairwise combination is "4/13". The embodiment can perform similarity judgment on long text operation passes, such as operation description characters in thousands of characters.
In some embodiments, the normalizing the surgical information to be processed by the classification algorithm of the present disclosure includes:
inputting the operation information to be processed into a neural network classification model;
analyzing whether the original operation information in the operation information to be processed corresponds to the standard words or not;
according to the analysis result, normalization processing is carried out on the corresponding original operation information and the standard words;
outputting standard surgical information corresponding to the original surgical information for generating the target medical data.
Specifically, continuing with the above example, assuming that the pair of pair is used as the surgical information to be processed, the present embodiment may use the pair of pair as the surgical information to be processed
"1. (original, non-drug eluting coronary stents placed)
2. (original, coronary artery drug coating stent implantation)
3. (original, drug eluting coronary stent)
……”
And inputting the neural network classification model. The method can construct an adaptive neural network classification model, and analyze and normalize the operation information to be processed in the modes of adaptively configuring corresponding frameworks, the number of neurons and the like.
For example, each group of information is classified, and it is determined whether the operation information from the original medical data and the standard word screened from the standard information table have related or similar medical concepts, for example, whether the two have a relationship of inclusion. If the neural network classification model considers that the two belong to the situation with the inclusion relationship, the neural network classification model outputs a corresponding result, for example, outputs "1" to characterize a positive conclusion. If the neural network classification model considers that the two do not belong to the situation with the inclusion relationship, the neural network classification model outputs a corresponding result, for example, outputs "0" to represent a negative conclusion.
In order to optimize the performance of the neural network classification model, in various embodiments of the present disclosure, the performance of the neural network classification model may be increased by optimizing a pre-training model. In particular, large-scale medical anticipation data may be collected, with the data sources targeting surgical information, including but not limited to: medical record data of each medical structure, medical data of each medical research institution, medical data in each medical information system, medical textbooks of various languages, medical dictionaries, medical treatises, medical treatment data in online databases, department libraries, website webpage data, forum data and the like. In the embodiment of the disclosure, a pre-training model can be constructed to pre-train the acquired large-scale medical prediction data, so that the model learns various medical knowledge from the large-scale data, and can be applied to specific tasks on the basis to improve the performance of the neural network model of the disclosure. Pre-training models include, but are not limited to: BERT model, XLNET model, roBERTa model, etc.
"1" for the aforementioned input neural network classification model based on the neural network classification model of the present embodiment (plain, non-drug eluting coronary stent placement)
2. (original, coronary artery drug coating stent implantation)
3. (original, drug eluting coronary stent)
… … ", standard surgical information corresponding to the original text may be output. Since the present disclosure can be aided with the ICD-9-CM-3 table, such standard surgical information may include standard surgical names, standard surgical name codes, etc., such as output:
"1.36.0700 | drug eluting coronary stent placement
2.00.6600X 004| percutaneous coronary artery saccule dilatation forming operation
3.00.4700 stent for three blood vessels
4.88.5700 | other and unspecified coronary angiography "
Intuitively, it can be understood that the above results outputted by the neural network classification model of the present embodiment can obtain accurate information description of the operation, including details about what type of stent is, how many stents are actually implanted, and how the patient has performed percutaneous coronary angioplasty.
As another example, an embodiment of the present disclosure provides an input of
"name of operation: temporary pacemaker Implantation permanent pacemaker Implantation (DDD)
The operation process comprises the following steps: the patient takes the horizontal position, and is disinfected and paved with 2 percent lidocaine local anesthesia by a conventional method. Puncturing the right femoral vein, feeding a guide steel wire along a puncture point, confirming that the guide steel wire is positioned in the inferior vena cava by an X-ray, feeding the guide steel wire into a sheath tube along the guide steel wire, pulling out the guide steel wire and an expansion sheath, feeding the temporary pacing electrode into a temporary pacing electrode through the sheath tube, feeding the electrode into the right ventricular apex part under the X-ray, connecting the pacing electrode and a pacemaker, measuring parameters, sensing pacing well, setting pacing voltage to be 5V, and setting the pacing voltage to be 40 times/minute. The skin is cut in parallel for about 5cm at the scar of the original operation in the left thoracic operation area, the subcutaneous tissue is separated layer by layer and blunt, the pacemaker is fully exposed and taken out, the size of the original bag is adjusted to be consistent with that of the new pacemaker, and the hemostatic gauze is fully filled for standby. The parameter threshold of the original atrial electrode is tested to be 0.8V, the impedance is 560 omega, and 3mv is sensed. The parameter threshold of the original ventricular electrode is tested to be 0.7V, the impedance is 480 omega, and 12mv is sensed. Order the patient to breathe cough deeply, the electrode does not have the dislocation, and the signal of pacing does not drop, and maximum voltage pacing diaphragm does not have the stimulation. Connecting the pulse generator and screwing the fixing screw. The pulse generator is placed into the pacemaker sac, and the pacemaker is fixed by stitching. Sewing the closed bag layer by layer, wrapping by local pressurization, and pressing with sand bag. The temporary pacing electrode catheter was withdrawn, the right inguinal area was bandaged, and the patient was ordered to brake for 6 hours on the right lower limb. The patient after the operation has nausea and vomit, the blood pressure is measured by 200/100mmHg, isosorbide dinitrate is given for quiescent point to control the blood pressure, the antiemetic treatment is performed, the symptom of the patient is relieved, the operation name is 'permanent pacemaker implantation', but the old pacemaker can be taken out and replaced by a new pacemaker according to the recorded content of the operation process and the medical knowledge, the replacement is the replacement of the pacemaker but not the implantation, and the replacement is the operation comprising the taking out and the re-implantation. Then, based on the data processing method of the present disclosure, at least the correction of the operation name according to the operation pass can be realized for the input, and the output can be the neural network classification model implemented by the present embodiment
"37.7800 | Implantation of temporary transvenous pacemaker System
37.8701| permanent pacemaker replacement surgery with dual chamber ".
Intuitively, it can be understood that the above results output by the neural network classification model of the present embodiment can obtain accurate information description of the operation, that is, the replacement of the pacemaker instead of the implantation, the replacement including the operations of extraction and re-implantation, and the like, when the old pacemaker is actually extracted and the new pacemaker is replaced.
In some embodiments, the neural network classification model of the present disclosure is obtained based on pre-training, and the pre-training mode includes:
acquiring original medical data containing original surgical information;
extracting operation name information and operation passing information in the original operation information;
combining the operation passing information with standard words in an operation name library, wherein the operation name library is constructed based on the operation name information and comprises an original operation name and a standard operation name of original operation information;
training a pre-training model by using the combined surgical procedure information and standard words in a surgical name library as training data;
and obtaining the neural network classification model based on the pre-training model.
In particular, embodiments of the present disclosure may construct and iteratively train a training model to which the present disclosure relates based on a sufficient amount of data to satisfy the training, such as one or more batches, or massive acquisitions of raw medical data containing raw surgical information.
Continuing with the foregoing example, the foregoing surgical information entitled "coronary angiography + PCI procedure" may be used as input data for the pre-trained model of the present embodiment. Based on the surgical name "coronarography + PCI surgery", the present embodiment can build a surgical name library that is intended to incorporate the original surgical name, and a standard surgical name as the primary purpose, to build data for training in combination with the surgical name. These data contain pairwise combinations of the original procedure and each standard word in the library of procedure names, each combination being the training data for training. On the basis of training the pre-training model of the present embodiment, the pre-training model can be used as an initialization model to train an end-to-end neural network classification model related to the embodiments of the present disclosure.
In some embodiments, the method of constructing the surgical name library of the present disclosure comprises:
inputting the operation name information into a neural network model for coding an operation name so as to extract a standard operation name code from the operation name information;
and constructing an operation name library containing the original operation names and the coded standard operation names based on the operation name information and the standard operation name codes.
Specifically, with the operation name "coronary angiography + PCI surgery", the operation name coding of the ICD standard can be extracted from the operation name text in the original medical data, preferably by means of a neural network model coding the operation name, in combination with the operation name "coronary angiography + PCI surgery", based on the result of the coding of the original operation name text
"1. other and unspecified coronary angiography
2. Non-drug eluting coronary stent placement "
The operation name library of the present embodiment is constructed, which includes the original operation name and the encoded standard operation name.
In the embodiment, the original operation name is combined with the standard word in the operation name library after the standard processing, so that a corresponding pre-training model is constructed, and an end-to-end neural network classification model is trained, thereby greatly improving the intelligent processing efficiency of the embodiment.
In particular to the neural network classification model in each embodiment of the present disclosure, if the pre-trained model in the embodiment of the present disclosure is used, the neural network classification model in the embodiment may be substantially consistent with the pre-trained model, and the pre-trained model is required to initialize the neural network classification model in the embodiment. The neural network classification model of the present embodiment may use any other classification model if the pre-trained model is not employed, including but not limited to: CNN, LSTM, transformer, etc. Based on the pre-trained model finetune in some embodiments, the model converges faster with a 2% improvement in performance.
In some embodiments, obtaining target medical data of the present disclosure includes:
characterizing the original surgical information in a standard surgical information format;
outputting the target medical data to update the raw medical data.
Specifically, with the above contents being combined, on the basis that the neural network classification model of the present embodiment outputs a corresponding result, and for a combination in which the output result is a positive conclusion, the present embodiment may extract the operation name and the operation code in the ICD-9-CM-3 standard table, and perform a unified expression process on the originally input operation data. Compared with the originally input operation data, the target medical data output by the embodiment can be used for accurately processing the original medical data in the modes of supplementing omission and correcting errors.
In some embodiments, the standard surgical information table of the present disclosure is constructed in a manner including:
and constructing based on the label of the ICD standard table, wherein the labeled result comprises the result according to the operation name and the coded result.
In combination with the foregoing, the present embodiment may construct a standard information table based on the labeling result, including information such as main codes, additional codes, operation names, and categories.
As one of the aspects of the present disclosure, as shown in fig. 2, the present disclosure also provides a processing apparatus of medical data, including:
an acquisition unit configured for acquiring raw medical data, the raw medical data containing raw surgical information;
the matching module is configured to match the original operation information with standard words of a standard operation information table to obtain a matching relation;
the screening module is configured to screen out to-be-processed surgical information respectively comprising original surgical information and standard words based on the matching relationship;
the processing module is configured to perform normalization processing on the surgical information to be processed through a classification algorithm to obtain target medical data.
In some embodiments, the obtaining unit of the present disclosure may be an input device, a screen capture device, a text recognition device, and the like, and is intended to enable obtaining a text generated based on an original input; and/or generating medical data based on AI algorithm recognition.
In combination with the foregoing, in some embodiments, the matching module of the present disclosure may be further configured to:
and combining the original operation information and each standard word of the standard operation information table in pairs, and respectively calculating the matching degree.
In combination with the foregoing, in some embodiments, the screening module of the present disclosure may be further configured to:
and under the condition that the matching degree meets a preset condition, taking the original operation information and the standard words of the standard operation information table as the operation information to be processed.
In combination with the foregoing, in some embodiments, the processing module of the present disclosure may be further configured to:
inputting the operation information to be processed into a neural network classification model;
analyzing whether the original operation information in the operation information to be processed corresponds to the standard words or not;
according to the analysis result, normalization processing is carried out on the corresponding original operation information and the standard words;
outputting standard surgical information corresponding to the original surgical information for generating the target medical data.
In some embodiments, the processing module of the present disclosure may be further configured to:
acquiring original medical data containing original surgical information;
extracting operation name information and operation passing information in the original operation information;
combining the operation passing information with standard words in an operation name library, wherein the operation name library is constructed based on the operation name information and comprises an original operation name and a standard operation name of original operation information;
training a pre-training model by using the combined surgical procedure information and standard words in a surgical name library as training data;
and obtaining the neural network classification model based on the pre-training model.
In some embodiments, the processing module of the present disclosure may be further configured to:
inputting the operation name information into a neural network model for coding an operation name so as to extract a standard operation name code from the operation name information;
and constructing an operation name library containing the original operation names and the coded standard operation names based on the operation name information and the standard operation name codes.
In particular, one of the inventive concepts of the present disclosure is directed to a method for medical procedure planning by obtaining raw medical data, the raw medical data containing raw surgical information; matching the original operation information with standard words of a standard operation information table to obtain a matching relation; screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship; and performing normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data, so that the target medical data can be generated based on a standard operation information table to process and sort the original medical data based on the original operation information, particularly the original medical data containing the operation process description content of the long text, to obtain a unified normalization processing result. According to the processing method, the idea and the method for extracting the standard operation name from the long text of the operation process can be applied to the scene of supplementing and correcting the original operation name by using the content of the operation process, the recall rate of the data processing performance from the upper end to the upper end reaches 95%, various operation information is completely, accurately and efficiently expressed by outputting the standard information and a uniform coding mode, and therefore the accuracy and the efficiency of medical research and medical diagnosis and treatment are improved.
As one of the aspects of the present disclosure, the present disclosure also provides a computer-readable storage medium having stored thereon computer-executable instructions, which when executed by a processor, mainly implement a processing method according to the medical data described above, including at least:
acquiring original medical data, wherein the original medical data comprises original operation information;
matching the original operation information with standard words of a standard operation information table to obtain a matching relation;
screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship;
and carrying out normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data.
In some embodiments, a processor executing computer-executable instructions may be a processing device including more than one general-purpose processing device, such as a microprocessor, Central Processing Unit (CPU), Graphics Processing Unit (GPU), or the like. More specifically, the processor may be a Complex Instruction Set Computing (CISC) microprocessor, Reduced Instruction Set Computing (RISC) microprocessor, Very Long Instruction Word (VLIW) microprocessor, processor running other instruction sets, or processors running a combination of instruction sets. The processor may also be one or more special-purpose processing devices such as an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), a system on a chip (SoC), or the like.
In some embodiments, the computer-readable storage medium may be a memory, such as a read-only memory (ROM), a random-access memory (RAM), a phase-change random-access memory (PRAM), a static random-access memory (SRAM), a dynamic random-access memory (DRAM), an electrically erasable programmable read-only memory (EEPROM), other types of random-access memory (RAM), a flash disk or other form of flash memory, a cache, a register, a static memory, a compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD) or other optical storage, a tape cartridge or other magnetic storage device, or any other potentially non-transitory medium that may be used to store information or instructions that may be accessed by a computer device, and so forth.
In some embodiments, the computer-executable instructions may be implemented as a plurality of program modules that collectively implement the method for displaying medical images according to any one of the present disclosure.
The present disclosure describes various operations or functions that may be implemented as or defined as software code or instructions. The display unit may be implemented as software code or modules of instructions stored on a memory, which when executed by a processor may implement the respective steps and methods.
Such content may be source code or differential code ("delta" or "patch" code) that may be executed directly ("object" or "executable" form). A software implementation of the embodiments described herein may be provided through an article of manufacture having code or instructions stored thereon, or through a method of operating a communication interface to transmit data through the communication interface. A machine or computer-readable storage medium may cause a machine to perform the functions or operations described, and includes any mechanism for storing information in a form accessible by a machine (e.g., a computing display device, an electronic system, etc.), such as recordable/non-recordable media (e.g., Read Only Memory (ROM), Random Access Memory (RAM), magnetic disk storage media, optical storage media, flash memory display devices, etc.). The communication interface includes any mechanism for interfacing with any of a hardwired, wireless, optical, etc. medium to communicate with other display devices, such as a memory bus interface, a processor bus interface, an internet connection, a disk controller, etc. The communication interface may be configured by providing configuration parameters and/or transmitting signals to prepare the communication interface to provide data signals describing the software content. The communication interface may be accessed by sending one or more commands or signals to the communication interface.
The computer-executable instructions of embodiments of the present disclosure may be organized into one or more computer-executable components or modules. Aspects of the disclosure may be implemented with any number and combination of such components or modules. For example, aspects of the disclosure are not limited to the specific computer-executable instructions or the specific components or modules illustrated in the figures and described herein. Other embodiments may include different computer-executable instructions or components having more or less functionality than illustrated and described herein.
The above description is intended to be illustrative and not restrictive. For example, the above-described examples (or one or more versions thereof) may be used in combination with each other. For example, other embodiments may be used by those of ordinary skill in the art upon reading the above description. In addition, in the foregoing detailed description, various features may be grouped together to streamline the disclosure. This should not be interpreted as an intention that a disclosed feature not claimed is essential to any claim. Rather, the subject matter of the present disclosure may lie in less than all features of a particular disclosed embodiment. Thus, the following claims are hereby incorporated into the detailed description as examples or embodiments, with each claim standing on its own as a separate embodiment, and it is contemplated that these embodiments may be combined with each other in various combinations or permutations. The scope of the disclosure should be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.
The above embodiments are merely exemplary embodiments of the present disclosure, which is not intended to limit the present disclosure, and the scope of the present disclosure is defined by the claims. Various modifications and equivalents of the disclosure may occur to those skilled in the art within the spirit and scope of the disclosure, and such modifications and equivalents are considered to be within the scope of the disclosure.
Claims (10)
1. A method of processing medical data, comprising:
acquiring original medical data, wherein the original medical data comprises original operation information;
matching the original operation information with standard words of a standard operation information table to obtain a matching relation;
screening out to-be-processed operation information respectively comprising original operation information and standard words based on the matching relationship;
and carrying out normalization processing on the operation information to be processed through a classification algorithm to obtain target medical data.
2. The method of claim 1, wherein,
the obtaining of the matching relationship comprises: combining the original operation information and each standard word of a standard operation information table in pairs, and respectively calculating matching degrees;
based on the matching relationship, screening out the to-be-processed operation information respectively containing the original operation information and the standard words, and the method comprises the following steps: and under the condition that the matching degree meets a preset condition, taking the original operation information and the standard words of the standard operation information table as the operation information to be processed.
3. The method of claim 1, wherein the normalizing the surgical information to be processed by the classification algorithm comprises:
inputting the operation information to be processed into a neural network classification model;
analyzing whether the original operation information in the operation information to be processed corresponds to the standard words or not;
according to the analysis result, normalization processing is carried out on the corresponding original operation information and the standard words;
outputting standard surgical information corresponding to the original surgical information for generating the target medical data.
4. The method of claim 3, wherein the neural network classification model is derived based on pre-training in a manner comprising:
acquiring original medical data containing original surgical information;
extracting operation name information and operation passing information in the original operation information;
combining the operation passing information with standard words in an operation name library, wherein the operation name library is constructed based on the operation name information and comprises an original operation name and a standard operation name of original operation information;
training a pre-training model by using the combined surgical procedure information and standard words in a surgical name library as training data;
and obtaining the neural network classification model based on the pre-training model.
5. The method of claim 4, wherein the library of procedure names is constructed in a manner comprising:
inputting the operation name information into a neural network model for coding an operation name so as to extract a standard operation name code from the operation name information;
and constructing an operation name library containing the original operation names and the coded standard operation names based on the operation name information and the standard operation name codes.
6. The method of claim 1, wherein obtaining target medical data comprises:
characterizing the original surgical information in a standard surgical information format;
outputting the target medical data to update the raw medical data.
7. The method of claim 1, wherein the standard surgical information table is constructed in a manner comprising:
and constructing based on the label of the ICD standard table, wherein the labeled result comprises the result according to the operation name and the coded result.
8. The method of claim 1, further comprising:
and analyzing the content of the original operation information to obtain an operation name and a long text operation process.
9. Apparatus for processing medical data, comprising:
an acquisition unit configured for acquiring raw medical data, the raw medical data containing raw surgical information;
the matching module is configured to match the original operation information with standard words of a standard operation information table to obtain a matching relation;
the screening module is configured to screen out to-be-processed surgical information respectively comprising original surgical information and standard words based on the matching relationship;
the processing module is configured to perform normalization processing on the surgical information to be processed through a classification algorithm to obtain target medical data.
10. A computer-readable storage medium having stored thereon computer-executable instructions that, when executed by a processor, implement:
the method of processing medical data according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011604300.2A CN112749307B (en) | 2020-12-30 | 2020-12-30 | Medical data processing method and device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011604300.2A CN112749307B (en) | 2020-12-30 | 2020-12-30 | Medical data processing method and device and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112749307A true CN112749307A (en) | 2021-05-04 |
| CN112749307B CN112749307B (en) | 2022-11-08 |
Family
ID=75647160
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011604300.2A Active CN112749307B (en) | 2020-12-30 | 2020-12-30 | Medical data processing method and device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112749307B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115689609A (en) * | 2022-10-28 | 2023-02-03 | 上海润益互联网科技有限公司 | Business opportunity pushing method, system, medium and equipment based on industrial information matching |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090070140A1 (en) * | 2007-08-03 | 2009-03-12 | A-Life Medical, Inc. | Visualizing the Documentation and Coding of Surgical Procedures |
| CN105069123A (en) * | 2015-08-13 | 2015-11-18 | 易保互联医疗信息科技（北京）有限公司 | Automatic coding method and system for Chinese surgical operation information |
| CN106919793A (en) * | 2017-02-24 | 2017-07-04 | 黑龙江特士信息技术有限公司 | A kind of data standardization processing method and device of medical big data |
| CN109256216A (en) * | 2018-08-14 | 2019-01-22 | 平安医疗健康管理股份有限公司 | Medical data processing method, device, computer equipment and storage medium |
| CN110767296A (en) * | 2019-10-09 | 2020-02-07 | 北京雅丁信息技术有限公司 | Operation coding method based on semantic similarity |
| CN111192667A (en) * | 2019-12-16 | 2020-05-22 | 山东众阳健康科技集团有限公司 | Method for prompting operation code based on intelligent operation |
| CN111191415A (en) * | 2019-12-16 | 2020-05-22 | 山东众阳健康科技集团有限公司 | Operation classification coding method based on original operation data |
-
2020
- 2020-12-30 CN CN202011604300.2A patent/CN112749307B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090070140A1 (en) * | 2007-08-03 | 2009-03-12 | A-Life Medical, Inc. | Visualizing the Documentation and Coding of Surgical Procedures |
| CN105069123A (en) * | 2015-08-13 | 2015-11-18 | 易保互联医疗信息科技（北京）有限公司 | Automatic coding method and system for Chinese surgical operation information |
| CN106919793A (en) * | 2017-02-24 | 2017-07-04 | 黑龙江特士信息技术有限公司 | A kind of data standardization processing method and device of medical big data |
| CN109256216A (en) * | 2018-08-14 | 2019-01-22 | 平安医疗健康管理股份有限公司 | Medical data processing method, device, computer equipment and storage medium |
| CN110767296A (en) * | 2019-10-09 | 2020-02-07 | 北京雅丁信息技术有限公司 | Operation coding method based on semantic similarity |
| CN111192667A (en) * | 2019-12-16 | 2020-05-22 | 山东众阳健康科技集团有限公司 | Method for prompting operation code based on intelligent operation |
| CN111191415A (en) * | 2019-12-16 | 2020-05-22 | 山东众阳健康科技集团有限公司 | Operation classification coding method based on original operation data |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115689609A (en) * | 2022-10-28 | 2023-02-03 | 上海润益互联网科技有限公司 | Business opportunity pushing method, system, medium and equipment based on industrial information matching |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112749307B (en) | 2022-11-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109920540A (en) | Construction method, device and the computer equipment of assisting in diagnosis and treatment decision system | |
| CN108366750A (en) | The system and method for ECG waveform adaptively interpreted | |
| CN107578798A (en) | The processing method and system of electronic health record | |
| CN107958708A (en) | Risk trend appraisal procedure and system after institute | |
| CN111317464B (en) | Electrocardiogram analysis method and device | |
| CN108320808A (en) | Analysis of medical record method and apparatus, equipment, computer readable storage medium | |
| CN112735544A (en) | Medical record data processing method and device and storage medium | |
| CN112749307B (en) | Medical data processing method and device and storage medium | |
| CN114186070A (en) | Medical term normalization method and device based on artificial intelligence model | |
| CN115295111A (en) | Structured report generation method, system and storage medium | |
| CN114913953A (en) | Medical entity relationship identification method and device, electronic equipment and storage medium | |
| CN110895580B (en) | ICD operation and operation code automatic matching method based on deep learning | |
| CN112542242A (en) | Data transformation/symptom scoring | |
| CN117497115A (en) | Automatic generation method, system, terminal and medium for medical information abstract | |
| CN112735545A (en) | Self-training method, model, processing method, device and storage medium | |
| CN112632910A (en) | Operation encoding method, electronic device and storage device | |
| CN110377698B (en) | Reading understanding-based task processing method and device, equipment and readable medium | |
| Nair et al. | Automated clinical concept-value pair extraction from discharge summary of pituitary adenoma patients | |
| US9408550B2 (en) | Evolving serial comparison system with critical alert notifications | |
| CN113392633B (en) | Medical named entity identification method, device and storage medium | |
| CN112700825B (en) | Medical data processing method, device and storage medium | |
| EP4226383A1 (en) | A system and a way to automatically monitor clinical trials - virtual monitor (vm) and a way to record medical history | |
| CN113434692A (en) | Method, system and equipment for constructing graph neural network model and recommending diagnosis and treatment scheme | |
| CN112989971A (en) | Electrocardiogram data fusion method and device for different data sources | |
| CN113140315A (en) | Health self-testing system, server and health detection system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |