36 results about "Main diagnosis" patented technology

The invention discloses an automatic filling method for a medical health record based on speech recognition. The method comprises that a training set file used for speech recognition training is obtained in advance; a speech recognition model of a deep neural network is called, and the model is trained to obtain an improved speech recognition model; a condition random field model is called and trained to obtain an information extraction model; voice data to be recognized is input, recognized by the improved speech recognition model automatically, and converted into text data; and feature of each word is obtained to form text feature data, the text feature data is marked, labels of each word in the text feature data is obtained, the words are logged into corresponding data according to the labels, and a record filled automatically is obtained. Compared with an automatic filling result in the prior art, objective facts and main diagnosis are combined, the reliability is very high, medical staff is effectively freed from tedious work of writing the health records of patients manually, the method is simple, and the accuracy is high.

The disclosure relates to a medicine-taking aiding decision-making method and device and belongs to the technical field of data processing. The method includes: clustering medical data according to patient basic information in medical data, determining the main diagnosis of the patient in each period according to discharge diagnosis in the medical data; determining effective medicine-taking data corresponding to the main diagnosis in each period according to medicine-taking data in the medical data, and clustering the multiple pieces of effective medicine-taking data to generate effective medicine-taking data clusters based on the main diagnosis; cleaning the effective medicine-taking data of the effective medicine-taking data clusters to determine a recommended medicine; acquiring the current main patient diagnosis, and determining and feeding back, in recommended medicine, medicine-taking data matching the current patient's main diagnosis. On the one hand, the method and device can accurately lock the main diagnosis, determine a corresponding relation between the diagnosed disease and the medicine-taking data, and on the other hand, can improve the accuracy of the effective medicine-taking data, thereby improving the accuracy of the evaluation of the medicine-taking effect, and can help the doctor expand the medicine-taking scope.

The invention provides a DRGs-based intelligent medical record home page filling and training system. The system comprises an information filling module, an information storage module, an informationoutput module, an information prompting module, a DRGs evaluation pre-judgment module and a medical record home page training module. The system can achieve automatic extraction and matching of medical record home page information, the system can acquire past similar medical record home page information and DRGs grouping information, prompt operation is performed according to a coding principle and a DRGs grouping scheme, a filling person can adjust main diagnosis and surgical operation codes according to the prompt to accurately classify the medical record into the correct DRGs group, the medical record home page training module can also provide training operation, students fill the medical record home page information according to the operation prompt, scoring is performed, and a simulated training function is achieved.

The invention provides a medical knowledge graph construction method and device, a medium and electronic equipment. The method comprises: acquiring a clinical data set, wherein the clinical data set comprises a plurality of doctor seeing records; calculating main diagnosis information in the treatment record, and performing data cleaning based on the main diagnosis information; extracting entitiesfrom the cleaned data, and determining entity relationship pairs and relationship weights among the entities in the entity relationship pairs according to at least two types of entities; and constructing a medical knowledge graph according to the entity relationship pair and the relationship weight between the entities in the entity relationship pair. According to the technical scheme, the flexibility of obtaining different entities is facilitated by setting the plurality of data sources, the entity information is further accurately obtained, and the accuracy of constructing the medical knowledge graph is further improved. Moreover, the process of constructing the medical knowledge graph supports cold start, and does not need to depend on the existing medical knowledge graph.

Embodiments of the invention provide a medical record text processing method and device, electronic equipment and a storage medium. The method comprises the steps of determining multiple candidate diagnosis types of a to-be-processed medical record text; determining the score of any candidate diagnosis type based on the correlation between any candidate diagnosis type and the medical record diagnosis and treatment information in the medical record text and the general diagnosis and treatment attribute of any candidate diagnosis type; and based on the score of each candidate diagnosis type, selecting a main diagnosis type of the medical record text from all the candidate diagnosis types. According to the method and device, the electronic equipment and the storage medium provided by the embodiment of the invention, the accuracy of disease diagnosis grouping is improved, and the processing efficiency of the medical record text is improved.

A system and method to aid diagnosis of a patient includes a data and knowledge acquisition module and a meta diagnosis prediction module. The data and knowledge acquisition module includes: an input for patient data including any of previous diagnosis, drugs, symptoms and treatment, an input for open data and an input for expert knowledge, and is arranged to use these inputs to create a patient clinical object (PCO), a biomedical knowledge graph and a rule based knowledge graph, and to enrich the patient clinical object using the biomedical knowledge graph. The meta diagnosis prediction module is arranged to use the patient clinical object and the biomedical knowledge graph and / or the rule based knowledge graph in two or more of the following predictors: a diagnosis-based predictor to provide a set of diagnoses based on previous diagnoses, a drug-based predictor to provide a set of diagnoses based on drugs taken by the patient, a symptom-based predictor to provide a set of diagnoses based on symptoms of the patient and a treatment-based predictor to provide a set of diagnoses based on the treatments the patient is receiving. The meta diagnosis prediction module includes a meta predictor to combine any of the sets of diagnoses to give a predicted primary diagnosis.

The invention discloses a DRG grouping method based on artificial intelligence, and the method comprises the steps: 1), obtaining medical record data and settlement data, automatically obtaining a data feature set through an AI technology, and carrying out the cutting of the obtained data; 2) dividing the data into 26 MDC according to the main diagnosis, and subdividing the 25 MDC into ADRG through combination with main surgical operation; 3) further subdividing the ADRG into three groups by combining CC and MCC on the basis of the ADRG; 4) selecting a core characteristic factor in the characteristic set, respectively clustering the subdivided three groups of medical record data of the ADRG by utilizing a clustering algorithm, and obtaining DRG; and 5) processing the clustered DRG by usinga decision forest, and combining partial groups to obtain a final DRG group. Disease diagnosis grouping DRG screening characteristic factors are automatically screened by using the artificial intelligence (AI) technology, and grouping is carried out in combination with clinical practice, so that the empirical constraints are broken through, and the grouping accuracy and effectiveness are improved.

The invention discloses an automatic grouping method for DRGs based on a convolutional neural network. The method comprises the following steps: collecting case data and grouping the data according tomain diagnosis categories and related grouping modes for core disease diagnosis; digitally encoding the data; constructing a shallow convolutional neural network model, clustering feature vectors extracted by the convolutional network by using a k-means clustering method to obtain k category labels, and performing iterative training through a combination of the category labels and a classifier supervision network; and after the model training is finished, performing data grouping application. By utilizing the method provided by the invention, the defects of manual feature selection and additional data annotation for newly-added grouping categories are avoided, and automatic learning grouping can be carried out on data which is fuzzy and difficult in grouping.

The invention discloses a normalization method and device for diagnosis words. The method comprises the following steps: acquiring a main diagnosis word and an attribute word of an original diagnosisword by structurally separating the original diagnosis word; determining whether the main diagnosis word and the attribute word of the original diagnosis word are the same as each main diagnosis wordand each attribute word of each standard diagnosis word in a standard word list or not; and normalizing the original diagnosis word to a standard diagnosis word when determining that the main diagnosis word and the attribute word of the original diagnosis word are both same as a main diagnosis word and an attribute word of the standard diagnosis word. With the normalization method and device for diagnosis words, input amount of manual labor is reduced and labor cost is reduced. The normalization efficiency of diagnosis words is enhanced. Meanwhile, the normalization accuracy of the original diagnosis word is improved.

The invention relates to a medical data processing method, a medical data processing device and a computer readable storage medium, and the processing method comprises the steps: obtaining original medical data which comprises an operation code and a diagnosis code; determining a target diagnosis code from diagnosis codes in the original medical data by taking the operation code as an index according to an operation diagnosis association table constructed based on the DRG grouping information; and generating target medical data by taking the diagnosis information corresponding to the target diagnosis code as main diagnosis information. The processing device includes: an acquisition unit, a processing module, and a generation module. According to the embodiments of the invention, the main diagnosis can be accurately and efficiently distinguished according to the operation code, so that the DRG grouping can be accurately carried out.

The invention belongs to the technical field of geoscience, and particularly relates to a borehole core alteration information recording method, which comprises the following steps of: 1, performing borehole core reflectivity spectral measurement; step 2, resampling spectral data; step 3, removing an envelope line; 4, calculating the absorption depth; step 5, carrying out data logging; step 6, carrying out visualization of a recording result. The hyperspectral technology advantages are utilized, the borehole core spectral data are rapidly obtained based on the portable spectrometer, and the borehole core alteration information is expressed in a semi-quantitative mode by calculating the alteration type main diagnosis spectral characteristic absorption depth.

The invention discloses a code recommendation method and system, corresponding equipment and a storage medium. The method comprises the steps: obtaining a medical record home page diagnosis text as the input of a trained first sub-eye inference model, the diagnosis text at least comprises a main diagnosis text, wherein the first sub-eye inference model comprises a deep learning language model anda sub-eye projection layer linked with the output of the deep learning language model, and the deep learning language model is used for at least outputting a main diagnosis semantic vector based on aninput diagnosis text; in response to the fact that the diagnosis text does not include the secondary diagnosis text, the sub-view projection layer is used for determining the sub-view classificationprobability of the corresponding primary diagnosis text for each international disease classification ICD based on the primary diagnosis semantic vector; and taking the m sub-items with the highest classification probability for the main diagnosis text as recommended main diagnosis ICD sub-items of the main diagnosis text. According to the invention, corresponding ICD codes can be intelligently and automatically recommended based on texts, and the accuracy and the efficiency are high.

The invention discloses a main diagnosis and main operation matching detection method and device, computing equipment and a storage medium. The method comprises the following steps: collecting and extracting a DIP directory base, preprocessing the extracted DIP directory base, and obtaining a DIP rule library containing an association pair {diagnosis code, operation code}; extracting and preprocessing diagnosis codes and operation codes in a medical record to construct a sample set; performing association rule extraction on the sample set by using an FP-growth algorithm, and extracting association pairs {diagnosis code, operation code} simultaneously having diagnosis codes and operation codes in 2-item set association rules to form an association rule base; integrating the DIP rule base and the association rule base to obtain a total rule base; and performing matching detection on the main diagnosis and the main operation in the medical record by using the general rule base. The quality of the home page of the medical record is improved and the workload of a coder is reduced by efficiently detecting the matching of the main diagnosis and the main operation.

The invention discloses a self-adaptive frequent itemset excavation method-based defect affinity analysis method and device. The method comprises the following steps: a data set for preset coding is obtained, and a total input data set is obtained; a frequent itemset of a first layer is generated; a frequent itemset is searched from a candidate set of each layer, and a corresponding support degreeis saved; a superset of an itemset of a corresponding layer is searched, and a candidate itemset of a next layer is generated; a scanning function is called to check whether the candidate itemset ofthe next layer has a frequent itemset, the candidate itemset of the next layer is built, the frequent itemset of each layer is stored in a knowledge database, and a key link relation is built via an equipment ID and a voltage transformer main diagnosis and evaluation system. Via the method, problems of a conventional method which is slow in static state data analysis and desynchrony between a uniform threshold value and a dynamic data scale can be effective solved via a self-adaptive threshold value algorithm; wide area data support-based voltage transformer defect affinity analysis can be realized, and precision of excavation results can be improved.

The invention discloses a drug adverse reaction processing method and a drug storage medium device. The drug adverse reaction processing method comprises the following steps that patient information and monitoring drug information in a hospital information system are acquired; the obtained medicine is stored in an intelligent medicine box, the intelligent medicine box comprises camera equipment, astorage chamber and a processor, the camera equipment sends shot and scanned information to the processor, and patient information and monitored medicine information are sent to the processor throughthe Internet; the intelligent medical box contacts an emergency contact person through a network according to pre-input emergency contact person information, is connected to a hospital information system through communication, establishes a private inquiry channel between a patient and a main diagnosis doctor, calls disease information of the patient, opens for the main diagnosis doctor, emergency medical guidance is provided for online medical treatment of the main diagnosis doctor through camera equipment. According to the invention, adverse reactions of drugs of patients can be detected, and more importantly, a solution can be provided rapidly, which is more beneficial to the health and safety of the patients.

The invention relates to a case grouping method and system. The method comprises the steps of obtaining case information, dividing the case information in corresponding basic groups according to main diagnosis codes and operation codes in the case information, and obtaining basic group codes and basic group names; when main diagnosis corresponding to the main diagnosis codes does not belong to a hospitalization time influence type, or, the basic groups do not belong to specific basic groups, performing calculation according to the basic group codes and the diagnosis codes, thereby obtaining diagnosis complexity scores corresponding to the diagnosis codes; according to the diagnosis complexity scores corresponding to the diagnosis codes, performing calculation to obtain a disease complexity index corresponding to the case information; and according to the disease complexity index, dividing the case information in detailed groups from the basic groups, and obtaining disease diagnosis related group codes, disease diagnosis related group names and disease diagnosis related group relative weights, thereby finishing case grouping. According to the case grouping method and system, unified encoding rules can be realized nationwide, so that unified case management of health administrative departments and medical insurance departments is facilitated; and the operability is strong.

The invention discloses a medical record home page main diagnosis prediction system and method. In the system, a data organization module extracts clinical data associated with a patient from a database, data cleaning and feature construction are conducted on the clinical data associated with the patient, and features associated with the patient are obtained; then, the crowd grouping module performs first-round grouping on the patients according to the patient-associated features to obtain patient-associated first-round grouping identifiers; and the model generation module completes training and generation of a corresponding main diagnosis prediction model for the patients in each first round group according to the associated features of the patients; finally, the result calculation engine module receives the main diagnosis prediction request, obtains a first round of grouping identifier associated with a newly-added patient through the data organization module and the crowd grouping module in sequence according to the newly-added patient identifier carried by the request, and calls the main diagnosis prediction model corresponding to the first round of grouping identifier to obtain a main diagnosis prediction result. By means of the system and method, the accuracy and the consistency of filling main diagnosis of the medical record home page can be improved.

The invention discloses a main diagnosis and main operation matching detection method and system based on federal mining, and the method comprises the steps: carrying out the mining of the cross combination of operation items and diagnosis items at each medical institution end based on the operation frequency and the diagnosis frequency, and uploading the related data of a mining result to a central node end; carrying out support degree, confidence degree and improvement degree analysis on the mining result related data at a central node end, and screening a matching rule pair of main diagnosis and main operation from the mining result related data according to an analysis result to construct a rule database; during application, clinical matching detection of the main diagnosis and the main operation is carried out by utilizing the matching rules in the rule database. According to the method and the system, mining and discovery of association rules are carried out by efficiently, safely and compliantly using massive medical record home page data of multiple medical institutions, so that the clinical matching between the main diagnosis and the main operation in the medical record home page is automatically detected.

The invention relates to a guided tumor-related main diagnosis selection method and system. The method comprises the following steps: presetting a plurality of statements, and trigger conditions between the statements and between the statements and a plurality of main diagnoses; and judging whether one of the trigger conditions is met or not according to an input signal, determining and outputting the next output statement, and repeatedly executing until one main diagnosis is determined from the plurality of main diagnoses. According to the technical scheme of the invention, by designing the statements, a coder can determine the main diagnosis in a targeted manner, does not need to memorize a tedious main diagnosis selection principle, and can automatically trigger the next statement only by answering the answers of the key points according to the content of the medical record, so that a guided main diagnosis selection strategy is realized and is easier to understand; and main diagnosis can be efficiently obtained.

The invention provides a grouping method and device, electronic equipment and a storage medium. The grouping device comprises: an obtaining module used for obtaining medical record home page data andexpense detail data, wherein the medical record home page data comprises a main diagnosis code, and the expense detail data comprises a first charging item code; a correction module which is used fordetermining a corrected main operation code based on the first charging item code, a preset charging item and operation relationship knowledge base, the main diagnosis code and a preset diagnosis andoperation knowledge base after judging that the main operation code of the medical record home page data needs to be corrected; and a grouping module which is used for grouping based on the main diagnosis code and the corrected main operation code to obtain a diagnosis related grouping DRG grouping result. According to the invention, grouping is carried out based on the corrected main operation codes, so that the accuracy of the DRG grouping result can be improved.

The invention belongs to the technical field of medical apparatuses, and discloses a handheld ultrasonic diagnosis apparatus. The handheld ultrasonic diagnosis apparatus comprises a main diagnosis apparatus body, a turnover lid, a transmission line and an ultrasonic probe, an electromagnetic device is arranged at one side of the main diagnosis apparatus body, and the main diagnosis apparatus bodyand the turnover lid are in bearing connection through bearings; the turnover lid is provided with a display screen, a permanent magnet is arranged at the top of the display screen, an operating panelis arranged at the top of the main diagnosis apparatus body, and a controller is arranged in the main diagnosis apparatus body, and electrically connected with the display screen. By means of the handheld ultrasonic diagnosis apparatus, power is supplied to the controller through a lithium battery, so that the controller controls the ultrasonic probe to work and examine a patient, information isfed back into the controller through the transmission line, the controller reflects the examination information on the display screen to give reference to a doctor, and the permanent magnet attracts an iron plate when the turnover lid is closed to protect the apparatus from being damaged; the handheld ultrasonic diagnosis apparatus is easy to operate and convenient to use and carry.

The invention discloses a clinical medication recommendation method based on multi-source heterogeneous data. The multi-source heterogeneous data comprises human information data including patient age, gender, ethnic group, education and main diagnosis information, wherein the main diagnosis information is regarded as static information, and the sequential examination result and the therapeutic drug are regarded as heterogeneous sequence data; according to the clinical medication intelligent recommendation method, obtained multi-source heterogeneous data is input into a model fusion network, and comprehensive patient representation information is obtained through learning by means of the deep learning technology.

The invention relates to a medical insurance fee control method and system based on an abnormity detection algorithm. The method comprises the steps: processing a medical record data text to obtain a vector mapping of an entity; according to the vector mapping of the entity, performing connotation consistency aggregation on all the main diagnoses in the case data, and dividing all the case data into N groups (N being greater than 1) according to the main diagnoses; and performing entity vector screening on the grouped cases, and running an LOF algorithm on each group of case data to obtain an abnormal medical behavior set. According to the method, the thought of big data is utilized, the abnormal medical behaviors in the medical process are found through the abnormity detection technology, and then medical insurance fee control can be carried out on the unreasonable behaviors.

A DRGs automatic grouping method based on convolutional neural network, including: collecting case data and grouping according to a major diagnostic broad categories and core diagnosis-related grouping method; performing numerical coding to the data; constructing a shallow convolutional neural network model, using a k-means clustering method to cluster the feature vectors extracted from the convolutional network to obtain k category labels, combining the category labels and classifier to supervise the network performing iterative training; after finishing training the model, perform data grouping application. The method of the present disclosure is used to avoid the disadvantages of manual feature selection and additional data labeling for adding new grouping categories, automatic learning grouping can be performed for data with vague and difficult groupings.

The invention provides a medical intelligent nutrition cup. The cup comprises a cup body, the cup body comprises a cup body shell and a cup body inner container. The cup body liner is arranged in thecup body shell; a cup body cavity is formed between the cup body liner and the cup body shell; a liquid metal piece and a microprocessor are arranged on the inner side wall of the cavity of the cup body; a power source and a weighing device are arranged on the inner bottom face of the cavity of the cup body, a fingerprint recognition device, a display screen and an alarm device are arranged on theside wall of the outer surface of the cup body shell, the power source is connected with a microprocessor, the display screen and the alarm device, and the microprocessor is connected with the alarmdevice, the weighing device, the display screen and the fingerprint recognition device. The cup is simple in structure, and convenient to operate. A patient makes contact with the fingerprint recognition device, the microprocessor obtains personal information of the patient, the microprocessor is communicated with the medical database and calls personal data of the patient, the personal data aredisplayed through the wireless display device, the patient, catering personnel, nutritionists and main diagnosis physicians can observe body shape information of the patient in real time, personalized catering of the patient is facilitated, and practicability is high.

The invention discloses a medical record statistical analysis system for complications and complications. The system comprises an analysis and calculation module for medical record statistical analysis; and a complication association query library, a medical record receiving system, a classification statistics module and an analysis module which communicate with the analysis and calculation module. According to the medical record statistical analysis system for complications and complications, whether main diagnosis results and other diagnosis results are complications or not, the complications and the probability of the complications can be completed, and prediction of one or more complications possibly caused by front-end diagnosis can be completed.

The invention discloses an online joint consultation method based on AI standardized auxiliary diagnosis and treatment decision in the technical field of working methods of AI standardized auxiliary diagnosis and treatment decision MDT. The method comprises the following steps: S1, a patient sees a doctor in a department, a medical record file is created through AI-MDT, a medical record picture and diagnosis and treatment image data are uploaded, and personalized diagnosis and treatment paths and schemes are provided through an AI diagnosis and treatment decision-making system according to the condition of the patient; and the AI standardized diagnosis and treatment decision-making system is introduced to be embedded into the consultation MDT process, main diagnosis doctors and consultation experts are assisted to give standardized diagnosis and treatment schemes, the system is easy to operate and convenient to use by the doctors. The problem that in the traditional consultation process, the doctors select treatment paths and schemes for patients according to own experience, and a certain probability of nonstandard treatment schemes can occur due to different doctor levels is solved. According to the system, the subsequent treatment path and treatment scheme of the patient can be calculated through AI (artificial intelligence) only by inputting the illness state of the patient.

A method for generating a medical report of a patient comprises the following steps that a main diagnosis doctor carries out training, then a specific two-dimensional code is generated for drainage, the patient pays attention to themain diagnosis doctor by scanning the two-dimensional code of the main diagnosis doctor, then doctor-patient binding is completed, and the main diagnosis doctor applies for the medical report through the case of the patient. The medical report of the patient is verified through an AI technology, then whether the medical report needs to be audited or not is determined, a final examination stage is started after an audit result is obtained, whether an application scheme of the medical report is consistent with an audit scheme or not is determined after final examination is completed, and a mode of pushing the medical report to the patient is selected; according to the method, the medical report is analyzed through the AI technology, and the accuracy of the medical report information can be improved through multi-layer verification and auditing; a patient scans a code through a mobile phone and carries out doctor-patient binding with a doctor, so that the medical report can be conveniently checked at a mobile phone terminal.