Medical examination consultation method and system based on medical big data
Through medical big data technology, voice feature information in medical testing consultation is obtained and processed, and the models are matched and corrected using big data comparison, which solves the error problems caused by relying on experience in the existing technology, and achieves more accurate inspection and evaluation information , avoid medical accidents.
Patent Information
- Application Number
- CN202411456364.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-18
- Publication Date
- 2024-11-29
AI Technical Summary
Existing medical testing consultations mainly rely on the experience of medical staff and are susceptible to subjective factors, resulting in errors and potential medical malpractice.
Using a method based on medical big data, the speech feature information is obtained, and the patient's condition feature vector is converted into a test feature vector and a disease feature vector, and the model is matched and corrected by big data comparison to provide more accurate test and evaluation information.
It improves the accuracy of inspection and evaluation information, reduces the influence of subjective factors, and avoids errors and medical accidents.
Smart Images

Figure CN119049622A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical data technology, and in particular to a medical examination consultation method and system based on medical big data. Background Art
[0002] Medical testing, also known as clinical testing or medical laboratory science, is a discipline that obtains information about an individual's health status through laboratory testing of samples taken from the human body (such as blood, urine, tissue samples, etc.), thereby providing an important basis for the prevention, diagnosis, treatment and later rehabilitation of diseases. Medical testing plays an important role in disease diagnosis and treatment. Medical testing consultation refers to a series of consulting services provided during the medical testing process, with the aim of helping doctors, patients or other healthcare providers to correctly understand and interpret medical test results, thereby making more accurate diagnosis and treatment decisions;
[0003] Existing test consultation mainly relies on the experience of medical staff. Therefore, it is easily affected by subjective factors, which in turn leads to certain errors in test consultation and thus medical accidents. Therefore, a medical test consultation method and system based on medical big data is needed to solve the above problems. Summary of the invention
[0004] The purpose of the present invention is to provide a medical examination consultation method and system based on medical big data to solve the technical problems raised in the above background technology.
[0005] To achieve the above object, the present invention provides the following technical solutions:
[0006] A medical test consultation method based on medical big data, comprising:
[0007] Acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information;
[0008] Acquire patient test information and patient condition information according to the patient inquiry information, convert the patient test information into a test feature vector, and then convert the patient condition information into a condition feature vector;
[0009] Performing text conversion on the doctor's reply information to obtain text information, and comparing the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information;
[0010] Converting the medical term feature information into a medical term feature vector, and then converting the inspection and evaluation information into an inspection and evaluation feature vector;
[0011] Input the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, and obtain the comparison medical test consultation information;
[0012] The test evaluation information is modified according to the comparison of the medical test consultation information to obtain the modified test evaluation information, and the patient is replied based on the modified test evaluation information.
[0013] Preferably, the step of acquiring patient test information and patient condition information according to the patient inquiry information, converting the patient test information into a test feature vector, and then converting the patient condition information into a condition feature vector comprises:
[0014] The patient inquiry information is disassembled according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format;
[0015] Comparing the digital information with a preset medical examination library to obtain examination digital information;
[0016] Screening the text information based on a preset medical word library to obtain test terms and patient information terms, and pairing the test terms with the test digital information based on a time series to obtain patient test information;
[0017] Acquire the patient's historical treatment information according to the patient inquiry information, acquire historical condition information according to the patient's historical treatment information, and acquire condition characteristic information according to the historical condition information;
[0018] Screening the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and using the patient condition characteristic information as the patient condition information;
[0019] The inspection digital information is spliced with a preset length based on the time series to obtain an inspection feature vector of a fixed length;
[0020] The patient's condition characteristic information is feature-splitting to obtain a plurality of condition characteristics, and the plurality of condition characteristics are spliced with a preset length based on a time series to obtain a condition characteristic vector.
[0021] Preferably, the step of comparing the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection and evaluation information further includes:
[0022] Cleaning the text information based on the Levenshtein distance algorithm to obtain cleaned text information;
[0023] Comparing the cleaned text information based on a preset medical word library to obtain medical terminology feature information and inspection evaluation information;
[0024] Acquire corresponding disease characteristic information according to the medical term characteristic information, associate the disease characteristic information with the test evaluation information, and obtain disease association information;
[0025] Compare the disease-associated information with the preset historical disease-associated examination and evaluation information to obtain difference information, obtain a difference ratio value according to the difference information and the preset historical disease-associated examination and evaluation information, and determine whether the difference ratio value exceeds the preset difference ratio value;
[0026] If exceeded, the disease characteristic information needs to be rechecked.
[0027] Preferably, the step of converting the medical term feature information into a medical term feature vector, and then converting the inspection and evaluation information into the inspection and evaluation feature vector comprises:
[0028] Acquire the occurrence frequency and inverse document frequency of the corresponding medical term according to the medical term feature information;
[0029] Acquire the total number of corresponding words according to the medical term feature information, and acquire the frequency occupancy weight and the inverse document frequency occupancy weight of the medical term according to the total number of words, the frequency of occurrence of the medical term and the inverse document frequency;
[0030] Converting the ratio of the frequency occupancy weight of the medical term to the inverse document frequency occupancy weight into a medical term feature vector;
[0031] Acquire corresponding inspection and evaluation vocabulary and test values according to the inspection and evaluation information, associate the inspection and evaluation vocabulary with the test values, and obtain inspection and evaluation associated data;
[0032] A corresponding correlation value is obtained according to the inspection and evaluation correlation data, and the correlation value is converted into an inspection and evaluation feature vector.
[0033] Preferably, the step of inputting the test feature vector, the condition feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching to obtain the compared medical test consultation information comprises:
[0034] Obtaining a standard test feature vector, a standard disease feature vector, a standard medical term feature vector, and a standard test evaluation feature vector corresponding to the test feature vector, the disease feature vector, the medical term feature vector, and the test evaluation feature vector;
[0035] The test feature vector, the disease feature vector, the medical term feature vector, the test evaluation feature vector, the standard test feature vector, the standard disease feature vector, the standard medical term feature vector and the standard test evaluation feature vector are input into the big data comparison model for matching to obtain the compared medical test consultation information, wherein the function formula of the big data comparison model is:
[0036]
[0037] Among them, B(d) represents the comparison value, j(y) represents the test feature vector, a represents the standard test feature vector, b(q) represents the disease feature vector, b represents the standard disease feature vector, y(l) represents the medical term feature vector, c represents the standard medical term feature vector, p(j) represents the test evaluation feature vector, and d represents the standard test evaluation feature vector;
[0038] The information corresponding to the comparison value is used as comparison medical test consultation information.
[0039] Preferably, the step of correcting the test evaluation information according to the comparison of the medical test consultation information to obtain the corrected test evaluation information, and responding to the patient based on the corrected test evaluation information includes:
[0040] Obtaining a standard evaluation value according to the compared medical test consultation information;
[0041] Acquire a corresponding first evaluation value according to the inspection and evaluation information;
[0042] Determining whether the first evaluation value is greater than a standard evaluation value;
[0043] If it is not greater than, it is determined that there is an error in the test evaluation information corresponding to the first evaluation value, and the test evaluation information is corrected based on the comparison of the medical test consultation information to obtain the corrected test evaluation information, and the patient is replied based on the corrected test evaluation information.
[0044] The present application also provides a medical examination consulting system based on medical big data, including:
[0045] A first acquisition module is used to acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information;
[0046] A second acquisition module is used to acquire patient test information and patient condition information according to the patient inquiry information, and convert the patient test information into a test feature vector, and then convert the patient condition information into a condition feature vector;
[0047] A first conversion module is used to perform text conversion on the doctor's reply information to obtain text information, and compare the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information;
[0048] A second conversion module is used to convert the medical term feature information into a medical term feature vector, and then convert the inspection and evaluation information into an inspection and evaluation feature vector;
[0049] The first comparison module is used to input the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, so as to obtain the comparison medical test consultation information;
[0050] The first correction module is used to correct the test evaluation information according to the comparison of the medical test consultation information, obtain the corrected test evaluation information, and respond to the patient based on the corrected test evaluation information.
[0051] Preferably, the second acquisition module includes:
[0052] A first disassembling unit is used to disassemble the patient inquiry information according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format;
[0053] A first comparison unit, used to compare the digital information with a preset medical test library to obtain test digital information;
[0054] A first screening unit is used to screen the text information based on a preset medical word library to obtain test terms and patient information terms, and to pair the test terms with the test digital information based on a time series to obtain patient test information;
[0055] A first acquisition unit is used to acquire the patient's historical treatment information according to the patient inquiry information, acquire historical condition information according to the patient's historical treatment information, and acquire condition characteristic information according to the historical condition information;
[0056] A second screening unit is used to screen the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and use the patient condition characteristic information as the patient condition information;
[0057] A first concatenation unit is used to concatenate the inspection digital information with a preset length based on the time series to obtain a fixed-length inspection feature vector;
[0058] The first splitting unit is used to perform feature splitting on the patient's condition characteristic information to obtain a plurality of condition characteristics, and splice the plurality of condition characteristics with a preset length based on a time series to obtain a condition characteristic vector.
[0059] The present application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.
[0060] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are implemented.
[0061] The beneficial effects of the present application are as follows: the present invention first obtains the patient's test information and the patient's condition information according to the patient's inquiry information, and converts the patient's test information into a test feature vector, and then converts the patient's condition information into a condition feature vector, and then performs text conversion on the doctor's reply information to obtain text information, and compares the text information based on a preset medical vocabulary to obtain medical terminology feature information and test evaluation information, and then converts the medical terminology feature information into a medical terminology feature vector, and then converts the test evaluation information into a test evaluation feature vector, and then inputs the test feature vector, the condition feature vector, the medical terminology feature vector and the test evaluation feature vector into a big data comparison model for matching to obtain compared medical test consultation information, and finally, the test evaluation information is corrected according to the compared medical test consultation information to obtain corrected test evaluation information, and the patient is replied based on the corrected test evaluation information, so that the corrected data is supplemented or corrected by historical data to make the test evaluation information more accurate, and at the same time will not be affected by subjective factors, which will cause certain errors in test consultation, thereby avoiding the problem of medical accidents caused by errors. BRIEF DESCRIPTION OF THE DRAWINGS
[0062] Figure 1 The figure is a schematic diagram of a method flow of an embodiment of the present application.
[0063] Figure 2 A schematic diagram of the system structure of an embodiment of the present application.
[0064] Figure 3 A schematic diagram of the internal structure of a computer device according to an embodiment of the present application.
[0065] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0066] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0067] like Figure 1-3 As shown, the present application provides a medical examination consultation method based on medical big data, comprising:
[0068] S1. Acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information;
[0069] S2, acquiring patient test information and patient condition information according to the patient inquiry information, and converting the patient test information into a test feature vector, and then converting the patient condition information into a condition feature vector;
[0070] S3, performing text conversion on the doctor's reply information to obtain text information, and comparing the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information;
[0071] S4, converting the medical term feature information into a medical term feature vector, and then converting the inspection and evaluation information into an inspection and evaluation feature vector;
[0072] S5, inputting the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, and obtaining the comparison medical test consultation information;
[0073] S6. Modify the test evaluation information according to the comparison of the medical test consultation information to obtain the modified test evaluation information, and respond to the patient based on the modified test evaluation information.
[0074] As described in the above steps S1-S6, the existing test consultation mainly relies on the experience of medical personnel, and is therefore easily affected by subjective factors, which in turn leads to certain errors in the test consultation, thereby causing medical accidents. Therefore, the present invention first obtains the voice feature information of the medical test consultation, wherein the voice feature information includes the patient's inquiry information and the doctor's reply information. In this way, after obtaining the voice feature information including the patient's inquiry information and the doctor's reply information, it can be determined whether there is an error based on the voice feature information including the patient's inquiry information and the doctor's reply information. At the same time, the data obtained in this way is real-time and comprehensive, and there is no omission problem. Then, the patient's test information and the patient's condition information are obtained according to the patient's inquiry information, and the patient's test information is converted into a test feature vector, and then the patient's condition information is converted into a condition feature vector. By quantifying and standardizing the patient's test information and the patient's condition information, a standard basis can be provided for data analysis and comparison. Then, the doctor's reply information is converted into text to obtain text information, and the text information is compared based on a preset medical vocabulary to obtain medical terminology feature information and test evaluation information. In this way, the oral feedback of doctors and patients can be converted into clear text information through text conversion, which is easy to understand and analyze. At the same time, by comparing with the preset medical vocabulary, the text information can be compared with the preset medical vocabulary to obtain medical terminology feature information and test evaluation information. After the vocabulary is compared, the obtained medical term feature information and inspection evaluation information can be made more accurate. Then, the medical term feature information is converted into a medical term feature vector, and then the inspection evaluation information is converted into an inspection evaluation feature vector. In this way, the medical term feature information and the inspection evaluation information can be converted into standardized numerical values for subsequent processing and analysis. Then, the inspection feature vector, the disease feature vector, the medical term feature vector and the inspection evaluation feature vector are used as important feature factors, and then the feature factors are compared with the preset standard values in the big data comparison model to obtain the compared medical inspection consultation information. In this way, the comparison results combined with historical data can help doctors make more accurate diagnoses and treatment recommendations. At the same time, automated big data comparison can greatly save doctors' time and improve work efficiency. It can also avoid errors in doctors' subjective judgments, which will produce certain differences after comparison. In this way, the inspection evaluation information can be corrected by comparing the medical inspection consultation information to obtain corrected inspection evaluation information, and the patient is replied based on the corrected inspection evaluation information. In this way, the corrected data is supplemented or corrected by historical data to make the inspection evaluation information more accurate, and at the same time, it will not be affected by subjective factors, which will cause certain errors in inspection consultation, thereby avoiding the problem of medical accidents caused by errors.
[0075] In one embodiment, the step S2 of acquiring patient test information and patient condition information according to the patient inquiry information, converting the patient test information into a test feature vector, and then converting the patient condition information into a condition feature vector includes:
[0076] S201, disassembling the patient inquiry information according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format;
[0077] S202, comparing the digital information with a preset medical examination library to obtain examination digital information;
[0078] S203, screening the text information based on a preset medical word library to obtain test terms and patient information terms, and pairing the test terms with the test digital information based on a time series to obtain patient test information;
[0079] S204, acquiring the patient's historical treatment information according to the patient inquiry information, acquiring historical condition information according to the patient's historical treatment information, and acquiring condition characteristic information according to the historical condition information;
[0080] S205, screening the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and using the patient condition characteristic information as the patient condition information;
[0081] S206, splicing the inspection digital information with a preset length based on the time series to obtain an inspection feature vector of a fixed length;
[0082] S207, feature-segmenting the patient's condition characteristic information to obtain a plurality of condition characteristics, and splicing the plurality of condition characteristics with a preset length based on a time series to obtain a condition characteristic vector.
[0083] As described in the above steps S201-S207, since there are certain differences in the information stated by different patients, the patient inquiry information is disassembled according to the preset format to obtain text information and digital information, wherein the preset format is a table editing format, which can make the patient inquiry information content more standardized, and then the digital information is compared with the preset medical test library to obtain the test digital information, wherein the digital information is the indicator parameter in the test process, for example: the hemoglobin test result is 161, the reference range is 130-175g / l, and the preset medical test library is the standard reference range, so after obtaining the patient test result, it can It is able to quickly judge the patient's condition and give a corresponding answer, so that after comparison with the preset medical test library, it can ensure that the digital information is associated with the reference range corresponding to the medical test to avoid misinterpretation. Then, the text information is screened based on the preset medical vocabulary to obtain test terms and patient information terms, and the test terms (hemoglobin in the above example) are paired with the test digital information based on the time series to obtain the patient's test information. In this way, the test terms and the test digital information are connected in series through the time series, so that the acquired data can be paired according to time without confusion. Then, the patient's historical treatment information is obtained based on the patient's inquiry information, and based on the patient's inquiry information, the patient's historical treatment information is obtained. The historical medical condition information is obtained according to the patient's historical treatment information, wherein the patient's historical treatment information is the case archive of the patient during the medical treatment process, and the condition characteristic information is obtained according to the historical medical condition information. The condition characteristic information in this case is the confirmed condition information of the patient during the previous treatment process. Therefore, using the confirmed condition information as an important feature for screening the patient's condition can quickly match the consultation purpose. The patient's treatment process and treatment background can be obtained through the condition characteristic information, and then under the longitudinal comparison of historical information, it is helpful to discover the trend of the condition, and at the same time, it can also be compared with the existing data to timely discover errors, and then the patient information entries are compared based on the condition characteristic information. Screening is performed to obtain the patient's condition characteristic information, and the patient's condition characteristic information is used as the patient's condition information. Screening through patient information entries can make the obtained data more concise and intuitive. Then, the test digital information is spliced with a preset length based on the time series to obtain a fixed-length test feature vector. The fixed-length test feature vector obtained in this way can be easily compared with the test results of other patients (or historical databases). The patient's condition characteristic information is feature-spliced to obtain multiple condition characteristics. Based on the time series, multiple condition characteristics are spliced with a preset length to obtain a condition characteristic vector. The standardized vector is convenient for subsequent training analysis.
[0084] In one embodiment, the step S3 of comparing the text information with a preset medical word library to obtain medical term feature information and inspection evaluation information further includes:
[0085] S301, cleaning the text information based on the Levenshtein distance algorithm to obtain cleaned text information;
[0086] S302, comparing the cleaned text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information;
[0087] S303, acquiring corresponding disease condition characteristic information according to the medical term characteristic information, and associating the disease condition characteristic information with the test evaluation information to obtain disease condition association information;
[0088] S304, comparing the disease association information with the preset historical disease inspection and evaluation information to obtain difference information, and obtaining a difference ratio value according to the difference information and the preset historical disease inspection and evaluation information, and determining whether the difference ratio value exceeds the preset difference ratio value;
[0089] If exceeded, the disease characteristic information needs to be rechecked.
[0090] As described in the above steps S301-S304, the present invention first cleans the text information based on the Levenshtein distance algorithm to obtain cleaned text information, so that the cleaned text information obtained can remove noise and correct spelling errors, and then compares the cleaned text information based on a preset medical vocabulary to obtain medical terminology feature information and test evaluation information. After such comparison, the medical terms in the doctor's reply can be accurately extracted, which helps the patient understand the doctor's professional opinion. Then, the corresponding disease characteristic information is obtained according to the medical terminology feature information, and the disease characteristic information is associated with the test evaluation information to obtain disease association information. Being able to integrate the test evaluation information with the disease characteristic information helps to form a more comprehensive description of the disease, and at the same time, it can also better understand the connection between the disease and the test results. Then, the disease-related information is compared with the test evaluation information of the preset historical disease to obtain the difference information, and the difference ratio value is obtained based on the difference information and the test evaluation information of the preset historical disease, and it is determined whether the difference ratio value exceeds the preset difference ratio value. If it exceeds, it is necessary to re-examine the disease characteristic information, so that abnormal conditions can be discovered in time through dynamic monitoring, and the changes in the disease can be monitored through difference analysis, so as to provide decision support for doctors and ensure the accuracy and safety of medical services. Such a processing flow not only improves the accuracy of information, but also enhances the analysis ability and security of the system, which helps to provide more accurate medical consultation and services.
[0091] In one embodiment, the step S4 of converting the medical term feature information into a medical term feature vector and then converting the inspection and evaluation information into the inspection and evaluation feature vector includes:
[0092] S401, acquiring the occurrence frequency and inverse document frequency of the corresponding medical term according to the medical term feature information;
[0093] S402, obtaining the total number of corresponding words according to the medical term feature information, and obtaining the frequency occupancy weight and the inverse document frequency occupancy weight of the medical term according to the total number of words, the frequency of occurrence of the medical term and the inverse document frequency;
[0094] S403, converting the ratio of the frequency occupancy weight of the medical term to the inverse document frequency occupancy weight into a medical term feature vector;
[0095] S404, acquiring corresponding inspection and evaluation vocabulary and test values according to the inspection and evaluation information, and associating the inspection and evaluation vocabulary with the test values to obtain inspection and evaluation associated data;
[0096] S405: Obtain corresponding correlation values according to the inspection and evaluation correlation data, and convert the correlation values into inspection and evaluation feature vectors.
[0097] As described in the above steps S401-S405, the present invention obtains the frequency of occurrence and the inverse document frequency of the corresponding medical term according to the medical term feature information, and by calculating the frequency and the inverse document frequency, the importance of each medical term can be measured, which helps to highlight the key terms, wherein the formula of the inverse document frequency is:
[0098] Among them, IDF(t) represents the inverse document frequency, N represents the total number of calculated texts, and n t Indicates the calculation frequency including t, 1+n tis not equal to zero, then the corresponding total number of words is obtained according to the medical term feature information, and the frequency occupancy weight and inverse document frequency occupancy weight of the medical term are obtained according to the total number of words, the frequency of occurrence of the medical term and the inverse document frequency. In this way, by calculating the frequency occupancy weight and the inverse document frequency occupancy weight, the importance of each term can be more accurately reflected. Then, the ratio of the frequency occupancy weight and the inverse document frequency occupancy weight of the medical term is converted into a medical term feature vector, and through the standardized feature vector representation, it can be easily compared and analyzed with other data, so that the feature vectorized data can be directly used as the input of the machine learning model. Then, according to the test evaluation information, the corresponding test evaluation vocabulary and test value are obtained, and the test evaluation vocabulary is associated with the test value to obtain the test evaluation associated data. Through the construction of the associated data, the test evaluation vocabulary and the test value can be closely linked, which is helpful to form more comprehensive test evaluation information. According to the test evaluation associated data, the corresponding associated value is obtained, and the associated value is converted into a test evaluation feature vector. By quantifying the data, high-quality data input can be provided for subsequent data analysis and model training.
[0099] In one embodiment, the step S5 of inputting the test feature vector, the condition feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching to obtain the compared medical test consultation information includes:
[0100] S501, acquiring a corresponding standard inspection feature vector according to the inspection feature vector, acquiring a corresponding standard condition feature vector according to the condition feature vector, acquiring a corresponding standard medical term feature vector according to the medical term feature vector, and acquiring a corresponding standard inspection evaluation feature vector according to the inspection evaluation feature vector;
[0101] S502, input the test feature vector, the disease feature vector, the medical term feature vector, the test evaluation feature vector, the standard test feature vector, the standard disease feature vector, the standard medical term feature vector and the standard test evaluation feature vector into the big data comparison model for matching to obtain a comparison value, wherein the function formula of the big data comparison model is:
[0102]
[0103] Among them, B(d) represents the comparison value, j(y) represents the test feature vector, a represents the standard test feature vector, b(q) represents the disease feature vector, b represents the standard disease feature vector, y(l) represents the medical term feature vector, c represents the standard medical term feature vector, p(j) represents the test evaluation feature vector, and d represents the standard test evaluation feature vector;
[0104] S503. Use the information corresponding to the comparison value as comparison medical test consultation information.
[0105] As described in the above steps S501-S503, the present invention first obtains the standard test feature vector, standard disease feature vector, standard medical terminology feature vector and standard test evaluation feature vector corresponding to the test feature vector, disease feature vector, medical terminology feature vector and test evaluation feature vector, and then inputs the test feature vector, disease feature vector, medical terminology feature vector, test evaluation feature vector, standard test feature vector, standard disease feature vector, standard medical terminology feature vector and standard test evaluation feature vector into the big data comparison model for matching to obtain the compared medical test consultation information, for example: test feature vector = [0.8, 0.6, 0.7], module length = 1.97, test feature vector = [0.9, 0.7, 0.8], module length =2.24, disease feature vector = [0.6, 0.5, 0.4], modulus = 0.94, test evaluation feature vector = [0.7, 0.8, 0.9], modulus = 1.87, while standard disease feature vector = [1, 1, 1], modulus = 1.73, standard medical term feature vector = = [1, 1, 1], modulus = 1.73, standard test evaluation feature vector = [1, 1, 1], modulus = 1.73, standard test evaluation feature vector = [1, 1, 1], modulus = 1.73, standard test evaluation feature vector = [1, 1, 1], modulus = 1.73, indicating a comparison value of 0.53. The comparison value obtained in this way can reflect the difference in data, while the standard value is 1. Therefore, the comparison value of 0.53 indicates that there are differences in the test evaluation information. The above data are only used for illustration and are not used as the only reference.
[0106] In one embodiment, the step S6 of correcting the test evaluation information according to the comparison of the medical test consultation information to obtain the corrected test evaluation information, and responding to the patient based on the corrected test evaluation information includes:
[0107] S601, obtaining a standard evaluation value according to the compared medical test consultation information;
[0108] S602, obtaining a corresponding first evaluation value according to the inspection and evaluation information;
[0109] S603, determining whether the first evaluation value is greater than a standard evaluation value;
[0110] If it is not greater than, it is determined that there is an error in the test evaluation information corresponding to the first evaluation value, and the test evaluation information is corrected based on the comparison of the medical test consultation information to obtain the corrected test evaluation information, and the patient is replied based on the corrected test evaluation information.
[0111] As described in the above steps S601-S603, after obtaining the comparison medical test consultation information, the present invention can determine in advance whether there is a difference in the detected data, and then after the advance judgment, it can obtain the standard judgment value according to the comparison medical test consultation information, and then obtain the corresponding first judgment value according to the test evaluation information, and then determine whether the first judgment value is greater than the standard judgment value;
[0112] If it is not greater than, it is determined that there is an error in the test evaluation information corresponding to the first judgment value, and the test evaluation information is corrected based on the comparison of the medical test consultation information to obtain the corrected test evaluation information, and the patient is replied based on the corrected test evaluation information. The corrected data is supplemented or corrected by historical data to make the test evaluation information more accurate, and at the same time it will not be affected by subjective factors, which will lead to certain errors in the test consultation, thereby avoiding the problem of medical accidents caused by errors.
[0113] The present application also provides a medical examination consulting system based on medical big data, including:
[0114] The first acquisition module 1 is used to acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information;
[0115] The second acquisition module 2 is used to acquire the patient's test information and the patient's condition information according to the patient's inquiry information, and convert the patient's test information into a test feature vector, and then convert the patient's condition information into a condition feature vector;
[0116] The first conversion module 3 is used to perform text conversion on the doctor's reply information to obtain text information, and compare the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information;
[0117] A second conversion module 4 is used to convert the medical term feature information into a medical term feature vector, and then convert the inspection and evaluation information into an inspection and evaluation feature vector;
[0118] The first comparison module 5 is used to input the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, and obtain the comparison medical test consultation information;
[0119] The first correction module 6 is used to correct the test evaluation information according to the comparison of the medical test consultation information, obtain the corrected test evaluation information, and respond to the patient based on the corrected test evaluation information.
[0120] In one embodiment, the second acquisition module includes:
[0121] A first disassembling unit is used to disassemble the patient inquiry information according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format;
[0122] A first comparison unit, used to compare the digital information with a preset medical test library to obtain test digital information;
[0123] A first screening unit is used to screen the text information based on a preset medical word library to obtain test terms and patient information terms, and to pair the test terms with the test digital information based on a time series to obtain patient test information;
[0124] A first acquisition unit is used to acquire the patient's historical treatment information according to the patient inquiry information, acquire historical condition information according to the patient's historical treatment information, and acquire condition characteristic information according to the historical condition information;
[0125] A second screening unit is used to screen the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and use the patient condition characteristic information as the patient condition information;
[0126] A first concatenation unit is used to concatenate the inspection digital information with a preset length based on the time series to obtain a fixed-length inspection feature vector;
[0127] The first splitting unit is used to perform feature splitting on the patient's condition characteristic information to obtain a plurality of condition characteristics, and splice the plurality of condition characteristics with a preset length based on a time series to obtain a condition characteristic vector.
[0128] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned medical examination consultation method based on medical big data.
[0129] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0130] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.
[0131] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A medical examination consultation method based on medical big data, It is characterized in that include: Acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information; Acquire patient test information and patient condition information according to the patient inquiry information, convert the patient test information into a test feature vector, and then convert the patient condition information into a condition feature vector; Performing text conversion on the doctor's reply information to obtain text information, and comparing the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information; Converting the medical term feature information into a medical term feature vector, and then converting the inspection and evaluation information into an inspection and evaluation feature vector; Input the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, and obtain the comparison medical test consultation information; The test evaluation information is modified according to the comparison of the medical test consultation information to obtain the modified test evaluation information, and the patient is replied based on the modified test evaluation information.
2. The medical examination consultation method based on medical big data according to claim 1, It is characterized in that The step of acquiring patient test information and patient condition information according to the patient inquiry information, converting the patient test information into a test feature vector, and then converting the patient condition information into a condition feature vector comprises: The patient inquiry information is disassembled according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format; Comparing the digital information with a preset medical examination library to obtain examination digital information; Screening the text information based on a preset medical word library to obtain test terms and patient information terms, and pairing the test terms with the test digital information based on a time series to obtain patient test information; Acquire the patient's historical treatment information according to the patient inquiry information, acquire historical condition information according to the patient's historical treatment information, and acquire condition characteristic information according to the historical condition information; Screening the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and using the patient condition characteristic information as the patient condition information; The inspection digital information is spliced with a preset length based on the time series to obtain an inspection feature vector of a fixed length; The patient's condition characteristic information is feature-splitting to obtain a plurality of condition characteristics, and the plurality of condition characteristics are spliced with a preset length based on a time series to obtain a condition characteristic vector.
3. The medical examination consultation method based on medical big data according to claim 1, It is characterized in that The step of comparing the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information also includes: Cleaning the text information based on the Levenshtein distance algorithm to obtain cleaned text information; Comparing the cleaned text information based on a preset medical word library to obtain medical terminology feature information and inspection evaluation information; Acquire corresponding disease characteristic information according to the medical term characteristic information, associate the disease characteristic information with the test evaluation information, and obtain disease association information; Compare the disease-associated information with the preset historical disease-associated examination and evaluation information to obtain difference information, obtain a difference ratio value according to the difference information and the preset historical disease-associated examination and evaluation information, and determine whether the difference ratio value exceeds the preset difference ratio value; If exceeded, the disease characteristic information needs to be rechecked.
4. The medical examination consultation method based on medical big data according to claim 1, It is characterized in that The step of converting the medical term feature information into a medical term feature vector, and then converting the inspection and evaluation information into an inspection and evaluation feature vector, comprises: Acquire the occurrence frequency and inverse document frequency of the corresponding medical term according to the medical term feature information; Acquire the total number of corresponding words according to the medical term feature information, and acquire the frequency occupancy weight and the inverse document frequency occupancy weight of the medical term according to the total number of words, the frequency of occurrence of the medical term and the inverse document frequency; Converting the ratio of the frequency occupancy weight of the medical term to the inverse document frequency occupancy weight into a medical term feature vector; Acquire corresponding inspection and evaluation vocabulary and test values according to the inspection and evaluation information, associate the inspection and evaluation vocabulary with the test values, and obtain inspection and evaluation associated data; A corresponding correlation value is obtained according to the inspection and evaluation correlation data, and the correlation value is converted into an inspection and evaluation feature vector.
5. The medical examination consultation method based on medical big data according to claim 1, It is characterized in that The step of inputting the test feature vector, the condition feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching to obtain the compared medical test consultation information includes: Obtaining a standard test feature vector, a standard disease feature vector, a standard medical term feature vector, and a standard test evaluation feature vector corresponding to the test feature vector, the disease feature vector, the medical term feature vector, and the test evaluation feature vector; The test feature vector, the disease feature vector, the medical term feature vector, the test evaluation feature vector, the standard test feature vector, the standard disease feature vector, the standard medical term feature vector and the standard test evaluation feature vector are input into the big data comparison model for matching to obtain the compared medical test consultation information, wherein the function formula of the big data comparison model is: Among them, B(d) represents the comparison value, j(y) represents the test feature vector, a represents the standard test feature vector, b(q) represents the disease feature vector, b represents the standard disease feature vector, y(l) represents the medical term feature vector, c represents the standard medical term feature vector, p(j) represents the test evaluation feature vector, and d represents the standard test evaluation feature vector; The information corresponding to the comparison value is used as comparison medical test consultation information.
6. The medical examination consultation method based on medical big data according to claim 1, It is characterized in that The step of correcting the test evaluation information according to the comparison of the medical test consultation information to obtain the corrected test evaluation information, and responding to the patient based on the corrected test evaluation information includes: Obtaining a standard evaluation value according to the compared medical test consultation information; Acquire a corresponding first evaluation value according to the inspection and evaluation information; Determining whether the first evaluation value is greater than a standard evaluation value; If it is not greater than, it is determined that there is an error in the test evaluation information corresponding to the first evaluation value, and the test evaluation information is corrected based on the comparison of the medical test consultation information to obtain the corrected test evaluation information, and the patient is replied based on the corrected test evaluation information.
7. A medical examination consulting system based on medical big data, It is characterized in that include: A first acquisition module is used to acquire voice feature information of medical examination consultation, wherein the voice feature information includes patient inquiry information and doctor response information; A second acquisition module is used to acquire patient test information and patient condition information according to the patient inquiry information, and convert the patient test information into a test feature vector, and then convert the patient condition information into a condition feature vector; A first conversion module is used to perform text conversion on the doctor's reply information to obtain text information, and compare the text information based on a preset medical vocabulary to obtain medical terminology feature information and inspection evaluation information; A second conversion module is used to convert the medical term feature information into a medical term feature vector, and then convert the inspection and evaluation information into an inspection and evaluation feature vector; The first comparison module is used to input the test feature vector, the disease feature vector, the medical term feature vector and the test evaluation feature vector into the big data comparison model for matching, so as to obtain the comparison medical test consultation information; The first correction module is used to correct the test evaluation information according to the comparison of the medical test consultation information, obtain the corrected test evaluation information, and respond to the patient based on the corrected test evaluation information.
8. The medical examination consulting system based on medical big data according to claim 7, It is characterized in that The second acquisition module includes: A first disassembling unit is used to disassemble the patient inquiry information according to a preset format to obtain text information and digital information, wherein the preset format is a table editing format; A first comparison unit, used to compare the digital information with a preset medical test library to obtain test digital information; A first screening unit is used to screen the text information based on a preset medical word library to obtain test terms and patient information terms, and to pair the test terms with the test digital information based on a time series to obtain patient test information; A first acquisition unit is used to acquire the patient's historical treatment information according to the patient inquiry information, acquire historical condition information according to the patient's historical treatment information, and acquire condition characteristic information according to the historical condition information; A second screening unit is used to screen the patient information entries based on the condition characteristic information to obtain the patient condition characteristic information, and use the patient condition characteristic information as the patient condition information; A first concatenation unit is used to concatenate the inspection digital information with a preset length based on the time series to obtain a fixed-length inspection feature vector; The first splitting unit is used to perform feature splitting on the patient's condition characteristic information to obtain a plurality of condition characteristics, and splice the plurality of condition characteristics with a preset length based on a time series to obtain a condition characteristic vector.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
Citation Information
Patent Citations
Medical examination information service system based on health records
CN109545335A
Method and device for processing clinical examination data, medium and terminal device
CN110600090A
Medical examination consultation system based on clinical examination medical big data
CN113972009A
Interrogation interaction method and device based on large language model and related equipment
CN117558439A
Method, device and equipment for generating medical data processing suggestions
CN118277532A
