WO2021139146A1 - 信息推荐方法、设备、计算机可读存储介质及装置 - Google Patents

信息推荐方法、设备、计算机可读存储介质及装置 Download PDF

Info

Publication number
WO2021139146A1
WO2021139146A1 PCT/CN2020/106345 CN2020106345W WO2021139146A1 WO 2021139146 A1 WO2021139146 A1 WO 2021139146A1 CN 2020106345 W CN2020106345 W CN 2020106345W WO 2021139146 A1 WO2021139146 A1 WO 2021139146A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
user
recommended
department
photo
Prior art date
Application number
PCT/CN2020/106345
Other languages
English (en)
French (fr)
Inventor
邹洪伟
Original Assignee
平安国际智慧城市科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安国际智慧城市科技股份有限公司 filed Critical 平安国际智慧城市科技股份有限公司
Publication of WO2021139146A1 publication Critical patent/WO2021139146A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/335—Filtering based on additional data, e.g. user or group profiles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/174—Facial expression recognition
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms

Definitions

  • This application relates to the technical field of artificial intelligence, and in particular to an information recommendation method, equipment, computer-readable storage medium and device.
  • hospitals mainly have hospital information systems (Hospital Information System, HIS), Laboratory Information Management System (Laboratory Information Management System, LIS), medical image archiving and communication system (Picture archiving and communication systems (PACS), radiology information management system (Radioiogy information system, RIS) and electronic medical records (Electronic Medical Record, EMR) and other information systems, the inventor found that these systems generally use patient medical record cards for authentication operations. Whenever you go to a place, you must first swipe your card to obtain patient-related information, and there are too many repeated operations. At present, hospital registration requires multi-level selection according to the default directory of the registration system to complete the registration operation. The registration process is cumbersome and time-consuming.
  • the main purpose of this application is to provide an information recommendation method, equipment, storage medium, and device, aiming to solve the technical problem of low registration efficiency in traditional Chinese hospitals in the prior art.
  • the information recommendation method includes the following operations:
  • the historical consultation department with the successful matching is taken as the first target department, and the first target department is recommended.
  • the information recommendation device includes a memory, a processor, and an information recommendation program stored on the memory and running on the processor.
  • the information recommendation program is configured to implement the following operations:
  • the historical consultation department with the successful matching is taken as the first target department, and the first target department is recommended.
  • this application also proposes a computer-readable storage medium on which an information recommendation program is stored, and when the information recommendation program is executed by a processor, the following operations are implemented:
  • the historical consultation department with the successful matching is taken as the first target department, and the first target department is recommended.
  • this application also proposes an information recommendation device, the information recommendation device including:
  • the obtaining module is configured to obtain the personal information of the user to be recommended, and to obtain the pre-examination information of the user to be recommended;
  • the search module is configured to search for patient medical data corresponding to the personal information from the hospital information system;
  • the classification module is configured to classify the medical data of the patient according to the medical department, and obtain the historical medical department of the user to be recommended and the medical keywords corresponding to each historical medical department;
  • the matching module is configured to perform word segmentation processing on the pre-questioning information, and matching all words obtained by performing word segmentation processing on the pre-questioning information with the medical keywords corresponding to each historical medical treatment department;
  • the recommendation module is configured to, after determining that the matching is successful, use the matched historical consultation department as the first target department, and recommend the first target department.
  • the patient medical data corresponding to the personal information of the user to be recommended is searched from the hospital information system, and the pre-examination information is obtained.
  • the pre-examination information is obtained.
  • the current patient’s physical condition information is combined.
  • FIG. 1 is a schematic structural diagram of an information recommendation device for a hardware operating environment involved in a solution of an embodiment of the present application
  • FIG. 2 is a schematic flowchart of the first embodiment of the application information recommendation method
  • FIG. 3 is a schematic flowchart of a second embodiment of an application information recommendation method
  • FIG. 4 is a schematic flowchart of a third embodiment of an application information recommendation method
  • FIG. 5 is a structural block diagram of the first embodiment of the information recommendation device of this application.
  • FIG. 1 is a schematic diagram of the structure of an information recommendation device for a hardware operating environment involved in a solution of an embodiment of the application.
  • the information recommendation device may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), communication bus 1002, user interface 1003, network interface 1004, and memory 1005.
  • a processor 1001 such as a central processing unit (Central Processing Unit, CPU)
  • communication bus 1002 is used to implement connection and communication between these components.
  • the user interface 1003 may include a display screen (Display), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.
  • the wired interface of the user interface 1003 may be a USB interface in this application.
  • the network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WI-FIdelity, WI-FI) interface).
  • WI-FIdelity wireless fidelity
  • the memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable memory (Non-volatile Memory, NVM), such as a disk memory.
  • RAM Random Access Memory
  • NVM Non-volatile Memory
  • the memory 1005 may also be a storage device independent of the aforementioned processor 1001.
  • FIG. 1 does not constitute a limitation on the information recommendation device, and may include more or less components than those shown in the figure, or combine certain components, or different component arrangements.
  • the memory 1005 which is a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and an information recommendation program.
  • the network interface 1004 is mainly used to connect to a back-end server and perform data communication with the back-end server; the user interface 1003 is mainly used to connect to user equipment; the information recommendation device is called by the processor 1001
  • the information recommendation program is stored in the memory 1005 and executes the information recommendation method provided in the embodiment of the present application.
  • FIG. 2 is a schematic flowchart of the first embodiment of the information recommendation method of this application, and the first embodiment of the information recommendation method of this application is proposed.
  • the information recommendation method includes the following operations:
  • Operation S10 obtain personal information of the user to be recommended, and obtain pre-examination information of the user to be recommended.
  • the execution subject of this embodiment is the information recommendation device, where the information recommendation device may be an electronic device such as a smart phone, a personal computer, or a server, which is not limited in this embodiment.
  • the personal information includes basic information such as name, age, ID number and gender.
  • the user to be recommended inputs the pre-inquiry information through a terminal.
  • the terminal may be a mobile phone of the user to be recommended, a hospital registration device, or the information recommendation device.
  • the current physical condition information or the information of the department that the user to be recommended enters through the information recommendation device may be selected through the options displayed on the display interface of the information recommendation device to generate the pre-interrogation information, and It may be that the user to be recommended speaks his current physical condition information by voice, specifically including: whether he has a fever, cough, runny nose, back pain, headache or eye pain, etc., and wants to go to medicine, dermatology, or ophthalmology, etc. Intentional department information.
  • Operation S20 Search for patient medical data corresponding to the personal information from the hospital information system.
  • the hospital information system will record the information of each patient's visit.
  • the patient medical data is the record information of the user to be recommended for historical medical treatment, including information such as historical medical departments, prescribed prescriptions, and basic medical conditions.
  • the personal information of each user and the corresponding patient medical data are recorded in the hospital information system, and the patient medical data corresponding to the personal information can be searched from the hospital information system.
  • Operation S30 classify the patient's medical data according to the medical department, and obtain the historical medical department of the to-be-recommended user and the medical keywords corresponding to each historical medical department.
  • the patient’s medical data can be classified according to the visiting department. Including dermatology, internal medicine, surgery, etc., and extract keywords such as information that can be treated by each historical clinic, and obtain the corresponding treatment information of each historical clinic to perform word segmentation processing, and obtain all words of the treatment disease information , Calculate the term frequency-inverse document frequency (Term Frequency-Inverse Document Frequency, TF-TDF) value. The larger the TF-TDF value, the more important the word. All words can be sorted according to the TF-TDF value from largest to smallest, and the preset number ranked first Words are used as keywords for medical consultation. The preset number can be set according to experience, for example, 3.
  • Operation S40 performing word segmentation processing on the pre-questioning information, and matching all words obtained by performing word segmentation processing on the pre-questioning information with the medical treatment keywords corresponding to each historical medical department.
  • the pre-inquiry information includes information such as the current physical condition of the user to be recommended or the clinic to be linked to.
  • the word segmentation processing on the pre-inquiry information can be performed through a dictionary-based word segmentation algorithm or It is string matching.
  • the string to be matched is matched with a sufficiently large dictionary based on a certain algorithm strategy. If the match hits, the word can be segmented. According to different matching strategies, it is divided into forward maximum matching method, reverse maximum matching method, two-way matching word segmentation, full segmentation path selection, etc., so as to obtain all the words of the pre-question information.
  • all words of the pre-interrogation information and the medical treatment keywords corresponding to each historical medical department are expressed in vector form, and all the words of the pre-interrogation information in the vector form are calculated to correspond to each historical medical department.
  • the cosine distance between the treatment keywords is used as the similarity. According to the similarity, it is judged whether all the words of the pre-questioning information match the treatment keywords corresponding to each historical treatment department.
  • the first similarity threshold can be preset by setting For example, 80%, when the similarity exceeds the preset first similarity threshold, the historical consultation department corresponding to the similarity that exceeds the preset first similarity threshold is identified as a department that has been successfully matched. It can also be recommended by selecting the department with the highest similarity in history as the department with a successful match.
  • Operation S50 After it is determined that the matching is successful, use the historically matched department as the first target department, and recommend the first target department.
  • the matching historical department is the department for which the user to be recommended needs to register, and then it is used as the first target department for recommendation, which may be the first target department in the information recommendation
  • the display interface of the device is displayed, and the to-be-recommended user confirms the first target department to complete the registration operation.
  • the first target department may also be played by voice, or the first target department may be recommended to the user terminal of the user to be recommended, such as a smart phone or smart watch, and the user to be recommended may be The first target department conducts follow-up registration operations.
  • the patient medical data corresponding to the personal information of the user to be recommended is searched from the hospital information system, and the pre-question information is obtained.
  • the pre-question information is obtained.
  • the current patient’s physical condition is obtained.
  • the information is combined with the medical data of the user’s medical history to recommend the user’s medical department more accurately, without requiring the user to select the multi-level options in the default directory in the registration system, directly and conveniently recommend the target department to the user, improving the efficiency of registration and improving user experience.
  • FIG. 3 is a schematic flowchart of a second embodiment of the information recommendation method of this application. Based on the first embodiment shown in FIG. 2 above, a second embodiment of the information recommendation method of this application is proposed.
  • the method further includes:
  • Operation S401 After determining that the matching fails, obtain other departments and corresponding other department keywords from the hospital information system except for the historical consultation department.
  • All words can be sorted according to the TF-TDF value from large to small, and the preset number of words ranked in the front can be obtained as the treatment Keywords, the preset number can be set according to experience, for example, 3, so as to obtain other department keywords corresponding to each other department.
  • Operation S402 match all words of the pre-questioning information with keywords of other departments corresponding to other departments.
  • a preset second similarity threshold can be set, such as 80%, when the similarity exceeds the preset When the second similarity threshold is used, other departments corresponding to the similarity that exceeds the preset second similarity threshold are identified as departments that have successfully matched. It can also be recommended by selecting other departments with the highest similarity as the departments with a successful match.
  • Operation S403 after it is determined that the matching is successful, another department that has been successfully matched is used as the second target department, and the second target department is recommended.
  • the other departments that are successfully matched are the departments that the user to be recommended need to register, and they are recommended as the second target department.
  • the second target department may be listed in the information recommendation device.
  • the display interface of the to-be-recommended user confirms the second target department, and the registration operation can be completed.
  • the second target department may also be played by voice, or the second target department may be recommended to the user terminal of the user to be recommended, such as a smart phone or smart watch.
  • the second target department performs follow-up registration operations.
  • Operation S404 After determining that the matching fails, send the pre-interrogation information to the target terminal, so that the medical staff can recommend a consultation department based on the pre-interrogation information through the target terminal.
  • the pre-inquiry information can be sent to the target terminal, which may be The personal computer or smart phone of the medical staff, so that the medical staff can view the pre-inquiry information through the target terminal, and recommend the medical department for the user to be recommended according to the pre-inquiry information, so as to increase the number of users to be recommended The efficiency of registration.
  • the target terminal which may be The personal computer or smart phone of the medical staff, so that the medical staff can view the pre-inquiry information through the target terminal, and recommend the medical department for the user to be recommended according to the pre-inquiry information, so as to increase the number of users to be recommended The efficiency of registration.
  • FIG. 4 is a schematic flowchart of a third embodiment of the information recommendation method of this application. Based on the above-mentioned first or second embodiment, a third embodiment of the information recommendation method of this application is proposed. This embodiment is described based on the first embodiment.
  • the method further includes:
  • Operation S101 Obtain a current photo of the user to be recommended.
  • the user to be recommended is usually a patient who goes to the hospital to see a doctor.
  • High-definition cameras are installed at the entrance of the building, the registration office or the entrance of each clinic, and the facial information and expression information of the patient are captured in real time.
  • the current photo of the user to be recommended is taken through a camera or other shooting equipment.
  • Operation S102 Perform feature extraction on the current photo to obtain the feature of the picture to be recognized corresponding to the current photo.
  • the current photo may be pre-processed, specifically, face detection and registration, face cutting, and image normalization.
  • a deep learning target detection algorithm based on candidate regions (Fast Region-based Convolutional Neural Networks, Fast-RCNN) to detect human faces, perform face cutting on the current photo, and obtain the face picture of the user to be recommended.
  • Image normalization is performed on the face picture, and the purpose of geometric normalization is mainly to transform the expression sub-images into a uniform size, which is beneficial to the extraction of expression features.
  • operation S102 includes:
  • the ginput(3) function is used to calibrate the three feature points of the eyes and nose, mainly by using the mouse to calibrate, and obtain the coordinate values of the three feature points. Then rotate the image according to the coordinate values of the left and right eyes to ensure the consistency of the face orientation.
  • the distance between the two eyes is d, and the midpoint is O.
  • Determine the rectangular feature area according to the facial feature points and the geometric model take O as the reference, cut d on the left and right sides, and take the rectangular areas of 0.5d and 1.5d in the vertical direction for cutting.
  • the scale transformation of the expression sub-region image into a uniform size is more conducive to the extraction of expression features.
  • the intercepted image is unified into a 90*100 image, and the geometric normalization of the image is realized, and the geometric normalized picture is obtained.
  • the gray scale normalization can also be performed on the geometrically normalized picture.
  • the gray scale normalization mainly increases the brightness of the image, makes the details of the image clearer, and reduces the influence of light and light intensity.
  • the feature extraction is performed on the gray-scale normalized image through the Principal Component Analysis (PCA) algorithm.
  • the principal component is the linear coefficient, that is, the projection direction, and the center of the coordinate axis is moved to the center of the data. , And then rotate the coordinate axis so that the variance of the data on the C1 axis is the largest, that is, the projections of all n data individuals in this direction are the most scattered, then more information is retained, C1 becomes the first principal component, C2
  • the second principal component find a C2, make the covariance (correlation coefficient) of C2 and C1 0, so as not to overlap with C1 information, and maximize the variance of the data in this direction, and so on, find the third principal component, The fourth principal component...
  • the p-th principal component There are p principal components for p random variables.
  • the feature value and feature vector are analyzed through covariance, and the feature vector is the feature face to obtain the feature of the image to be recognized. Based on dynamic pictures, geometric methods or deep learning methods can also be used.
  • Operation S103 Perform micro-expression detection according to the feature of the picture to be recognized, and obtain the current expression of the user to be recommended.
  • micro-expression detection is used to predict the patient's mental state, such as whether the patient is happy, uncomfortable, angry, or sad.
  • Hospital staff will treat different patients with different attitudes according to the patient's psychological state to reduce doctor-patient disputes. For example: When the patient is in an angry state of mind, the medical staff should try to use a gentle tone to avoid irritating the patient and causing medical injuries; when the patient is in a sad state of mind, the medical staff should try not to despise the patient and say more encouragement Words, to prevent the patient from being more sad, or even triggering suicide.
  • operation S103 includes:
  • micro-expression detection uses deep learning through a multi-layer network structure, such as a recurrent neural network (Recurrent Neuron).
  • Networks RNNs
  • RNNs recurrent neural network
  • Deep neural networks can recognize facial expressions end-to-end.
  • One way is to add a loss layer at the end of the network to correct the back propagation error. The predicted probability of each sample can be directly output from the network.
  • Another way is to use a deep neural network as a tool to extract features, and then use a traditional classifier, such as a random forest model, to classify the extracted features, so as to obtain the current expression of the user to be recommended.
  • Operation S104 Find the corresponding service attitude suggestion according to the current expression, and send the service attitude suggestion to the target terminal.
  • the current facial expressions of the user to be recommended are detected, and the current facial expressions include happy, uncomfortable, angry, sad, impatient, and so on.
  • Corresponding service attitude suggestions can be preset for various expressions. For example, when the current expression is angry, the corresponding service attitude suggestions are: use a gentle tone to avoid irritating the patient.
  • the target terminal may be a smart phone or a computer of a medical staff, so that the medical staff can understand the mood of the patient in time and find a suitable way to communicate, so as to improve the efficiency of treatment.
  • the obtaining the personal information of the user to be recommended includes:
  • the personal information of the target user is obtained from the public security system database as the personal information of the user to be recommended.
  • the photo features corresponding to each photo can be used to align the face to the average face, so that the positions of the face feature points in all images are almost the same after the alignment.
  • the face recognition algorithm trained with aligned images is more effective. Perform facial feature point positioning on the image features to be recognized to obtain the face feature points to be processed corresponding to the image features to be recognized; compare the face feature points to be processed with preset facial feature points to obtain a homography matrix; The homography matrix transforms the face in the photo to obtain a calibrated face picture; compares the calibrated face picture with each photo feature in the security system library through a convolutional neural network model to obtain the to-be-recognized The similarity of the face between the image feature and each photo feature.
  • the preset face similarity threshold may be set according to an empirical value, such as 80%.
  • the information recommendation method further includes:
  • the personal information of the user to be recommended is obtained through the public security system database through face recognition. If the user to be recommended has a criminal record, it is determined that the user to be recommended is a doctor-patient risk object, and the The judgment result is sent to the target terminal.
  • the target terminal may be a smartphone or computer of a medical staff to notify the security staff and medical staff to pay attention to prevent medical injuries.
  • the micro-expression detects that the patient is in a pre-injury and medical condition, it will proactively notify the security and public security organs, let them come to protect and stop, and notify the medical staff to take protective measures to ensure their own safety and the safety of hospital property.
  • the collected data can be submitted to management agencies and judicial organs to crack down on medical personnel with reasonable evidence.
  • the current expression of the user to be recommended searches for the corresponding service attitude suggestion based on the current expression, and sends the service attitude suggestion to the target terminal so that the medical staff can understand the patient’s mood in time and find a suitable way to communicate. Improve the efficiency of treatment.
  • an embodiment of the present application also proposes a computer-readable storage medium having an information recommendation program stored on the computer-readable storage medium, and when the information recommendation program is executed by a processor, the information recommendation method described above is implemented. operating.
  • the computer-readable storage medium may be non-volatile or volatile, and the computer-readable storage medium may be referred to as "storage medium" for short.
  • an embodiment of the present application also proposes an information recommendation device, and the information recommendation device includes:
  • the obtaining module 10 is configured to obtain the personal information of the user to be recommended, and to obtain the pre-examination information of the user to be recommended.
  • the personal information includes basic information such as name, age, ID number and gender.
  • the user to be recommended inputs the pre-inquiry information through a terminal, and the terminal may be a mobile phone of the user to be recommended, a hospital registration device, or the information recommendation device.
  • the current physical condition information or the information of the department that the user to be recommended enters through the information recommendation device may be selected through the options displayed on the display interface of the information recommendation device to generate the pre-interrogation information, and It may be that the user to be recommended speaks his current physical condition information through voice, specifically including: whether he has fever, cough, runny nose, low back pain, headache or eye pain, etc., and wants to go to medicine, dermatology or ophthalmology, etc. Intentional department information.
  • the searching module 20 is configured to search for patient medical data corresponding to the personal information from the hospital information system.
  • the hospital information system will record the information of each patient's visit.
  • the patient medical data is the record information of the user to be recommended for historical medical treatment, including information such as historical medical departments, prescribed prescriptions, and basic medical conditions.
  • the personal information of each user and the corresponding patient medical data are recorded in the hospital information system, and the patient medical data corresponding to the personal information can be searched from the hospital information system.
  • the classification module 30 is configured to classify the medical data of the patient according to the medical department, and obtain the historical medical department of the user to be recommended and the medical keywords corresponding to each historical medical department.
  • the patient’s medical data can be classified according to the visiting department. Including dermatology, internal medicine, surgery, etc., and extract keywords such as information that can be treated by each historical clinic, and obtain the corresponding treatment information of each historical clinic to perform word segmentation processing, and obtain all words of the treatment disease information , Calculate the term frequency-inverse document frequency (Term Frequency-Inverse Document Frequency, TF-TDF) value. The larger the TF-TDF value, the more important the word. All words can be sorted according to the TF-TDF value from largest to smallest, and the preset number ranked first Words are used as keywords for medical consultation. The preset number can be set based on experience, for example, 3.
  • the matching module 40 is configured to perform word segmentation processing on the pre-questioning information, and matching all words obtained by performing word segmentation processing on the pre-questioning information with the medical treatment keywords corresponding to each historical medical department.
  • the pre-inquiry information includes information such as the current physical condition of the user to be recommended or the clinic to be linked to.
  • the word segmentation processing on the pre-inquiry information can be performed through a dictionary-based word segmentation algorithm or It is string matching.
  • the string to be matched is matched with a sufficiently large dictionary based on a certain algorithm strategy. If the match hits, the word can be segmented. According to different matching strategies, it is divided into forward maximum matching method, reverse maximum matching method, two-way matching word segmentation, full segmentation path selection, etc., so as to obtain all the words of the pre-question information.
  • all words of the pre-interrogation information and the medical treatment keywords corresponding to each historical medical department are expressed in vector form, and all the words of the pre-interrogation information in the vector form are calculated to correspond to each historical medical department.
  • the cosine distance between the treatment keywords is used as the similarity. According to the similarity, it is judged whether all the words of the pre-questioning information match the treatment keywords corresponding to each historical treatment department.
  • the first similarity threshold can be preset by setting For example, 80%, when the similarity exceeds the preset first similarity threshold, the historical consultation department corresponding to the similarity that exceeds the preset first similarity threshold is identified as a department that has been successfully matched. It can also be recommended by selecting the department with the highest similarity in history as the department with a successful match.
  • the recommendation module 50 is configured to, after determining that the matching is successful, use the matched historical consultation department as the first target department, and recommend the first target department.
  • the matching historical department is the department for which the user to be recommended needs to register, and then it is used as the first target department for recommendation, which may be the first target department in the information recommendation
  • the display interface of the device is displayed, and the to-be-recommended user confirms the first target department to complete the registration operation.
  • the first target department may also be played by voice, or the first target department may be recommended to the user terminal of the user to be recommended, such as a smart phone or smart watch, and the user to be recommended may be The first target department conducts follow-up registration operations.
  • the patient medical data corresponding to the personal information of the user to be recommended is searched from the hospital information system, and the pre-question information is obtained.
  • the pre-question information is obtained.
  • the current patient’s physical condition is obtained.
  • the information is combined with the medical data of the user’s medical history to recommend the user’s medical department more accurately, without requiring the user to select the multi-level options in the default directory in the registration system, directly and conveniently recommend the target department to the user, improving the efficiency of registration and improving user experience.
  • the information recommendation device further includes:
  • the obtaining module 10 is further configured to obtain other departments and corresponding keywords of other departments from the hospital information system after determining that the matching fails;
  • the matching module 40 is further configured to match all words of the pre-questioning information with keywords of other departments corresponding to other departments;
  • the recommendation module 50 is further configured to, after determining that the matching is successful, use other departments that are successfully matched as the second target department, and recommend the second target department;
  • the sending module is configured to send the pre-interrogation information to a target terminal after determining that the matching fails, so that the medical staff can recommend a treatment department based on the pre-interrogation information through the target terminal.
  • the information recommendation device further includes:
  • the obtaining module 10 is further configured to obtain the current photo of the user to be recommended;
  • the feature extraction module is configured to perform feature extraction on the current photo to obtain the feature of the picture to be recognized corresponding to the current photo;
  • the micro-expression detection module is configured to perform micro-expression detection according to the characteristics of the picture to be recognized, and obtain the current expression of the user to be recommended;
  • the search module 20 is further configured to search for a corresponding service attitude suggestion according to the current expression, and send the service attitude suggestion to the target terminal.
  • the feature extraction module is configured to perform geometric normalization on the current photo to obtain a geometrically normalized picture; perform gray-scale normalization on the geometrically normalized picture to obtain a grayscale A normalized picture; feature extraction is performed on the gray-scale normalized picture through a principal component analysis algorithm to obtain the characteristics of the picture to be recognized corresponding to the current photo.
  • the classification module 30 is further configured to learn the features of the image to be recognized through a recurrent neural network model, and to classify the learned features through a random forest model, to obtain the current profile of the user to be recommended expression.
  • the acquisition module 10 is further configured to acquire a photo collection from a public security system library, perform feature extraction on each photo in the photo collection, and obtain photo features corresponding to each photo; and compare the features of the pictures to be identified with Each photo feature is matched, and after determining that the matching is successful, the target user corresponding to the successfully matched photo is identified as the user to be recommended; the personal information of the target user is obtained from the public security system database as the personal information of the user to be recommended .
  • the information recommendation device further includes:
  • a judging module configured to judge whether the user to be recommended is a doctor-patient risk object according to the personal information of the user to be recommended, and obtain a judgment result
  • the sending module is further configured to send the judgment result to the target terminal.
  • Memory image ROM/Random Access Memory (Random Access Memory, RAM, magnetic disk, optical disk), including several instructions to make a terminal device (can be a mobile phone, computer, server, air conditioner, or network equipment, etc.) ) Perform the methods described in each embodiment of the present application.
  • ROM Read Only Memory
  • RAM Random Access Memory
  • magnetic disk magnetic disk
  • optical disk including several instructions to make a terminal device (can be a mobile phone, computer, server, air conditioner, or network equipment, etc.) ) Perform the methods described in each embodiment of the present application.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • General Business, Economics & Management (AREA)
  • Library & Information Science (AREA)
  • Epidemiology (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Business, Economics & Management (AREA)
  • Computational Linguistics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本申请公开了一种信息推荐方法、设备、存储介质及装置，该方法通过获取待推荐用户的个人信息及预问诊信息，从医院信息系统中查找与个人信息对应的患者医疗数据，对患者医疗数据按照就诊科室进行分类，获得待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词，将对预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配，将匹配成功的历史就诊科室进行推荐。

Description

信息推荐方法、设备、计算机可读存储介质及装置
本申请要求于2020年1月9日提交中国专利局、申请号为202010024389.9，发明名称为“信息推荐方法、设备、存储介质及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及人工智能的技术领域，尤其涉及一种信息推荐方法、设备、计算机可读存储介质及装置。
 背景技术
现阶段医院主要有医院信息系统(Hospital Information System，HIS)、实验室信息管理系统(Laboratory Information Management System，LIS)、医学影像存档与通讯系统(Picture archiving and communication systems，PACS)、放射信息管理系统(Radioiogy information system，RIS)和电子病历 (Electronic Medical Record，EMR)等信息化系统，发明人发现，这些系统一般通过患者病历卡进行认证操作，每到一个地方要先刷卡才能获得病人相关信息，重复操作太多。目前，医院挂号需按照挂号系统的默认目录进行多层级选择，才能完成挂号操作，挂号流程繁琐、耗时长。
上述内容仅用于辅助理解本申请的技术方案，并不代表承认上述内容是现有技术。
 技术解决方案
本申请的主要目的在于提供一种信息推荐方法、设备、存储介质及装置，旨在解决现有技术中医院挂号效率低的技术问题。
为实现上述目的，本申请提供一种信息推荐方法，所述信息推荐方法包括以下操作：
获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
从医院信息系统中查找与所述个人信息对应的患者医疗数据；
对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
此外，为实现上述目的，本申请还提出一种信息推荐设备，所述信息推荐设备包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的信息推荐程序，所述信息推荐程序配置为实现如下操作：
获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
从医院信息系统中查找与所述个人信息对应的患者医疗数据；
对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
此外，为实现上述目的，本申请还提出一种计算机可读存储介质，所述计算机可读存储介质上存储有信息推荐程序，所述信息推荐程序被处理器执行时实现如下操作：
获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
从医院信息系统中查找与所述个人信息对应的患者医疗数据；
对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
此外，为实现上述目的，本申请还提出一种信息推荐装置，所述信息推荐装置包括：
获取模块，配置为获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
查找模块，配置为从医院信息系统中查找与所述个人信息对应的患者医疗数据；
分类模块，配置为对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
匹配模块，配置为对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
推荐模块，配置为确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
本申请中，从医院信息系统中查找与待推荐用户的个人信息对应的患者医疗数据，并获取预问诊信息，通过将预问诊信息与患者医疗数据进行匹配，即将当前患者身体状况信息结合用户历史就诊的患者医疗数据，更准确地为用户推荐就诊科室，无需用户对挂号系统中默认目录的多层级选项进行逐层选择，直接便捷地为用户推荐目标科室，提高挂号效率，提升用户体验。
 附图说明
图1是本申请实施例方案涉及的硬件运行环境的信息推荐设备的结构示意图；
图2为本申请信息推荐方法第一实施例的流程示意图；
图3为本申请信息推荐方法第二实施例的流程示意图；
图4为本申请信息推荐方法第三实施例的流程示意图；
图5为本申请信息推荐装置第一实施例的结构框图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
本申请的实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
参照图1，图1为本申请实施例方案涉及的硬件运行环境的信息推荐设备结构示意图。
如图1所示，该信息推荐设备可以包括：处理器1001，例如中央处理器（Central Processing Unit，CPU），通信总线1002、用户接口1003，网络接口1004，存储器1005。其中，通信总线1002用于实现这些组件之间的连接通信。用户接口1003可以包括显示屏（Display），可选用户接口1003还可以包括标准的有线接口、无线接口，对于用户接口1003的有线接口在本申请中可为USB接口。网络接口1004可选的可以包括标准的有线接口、无线接口（如无线保真（WIreless-FIdelity，WI-FI）接口）。存储器1005可以是高速的随机存取存储器（Random Access Memory，RAM）存储器，也可以是稳定的存储器（Non-volatile Memory，NVM），例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储装置。
本领域技术人员可以理解，图1中示出的结构并不构成对信息推荐设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
如图1所示，作为一种计算机可读存储介质的存储器1005中可以包括操作系统、网络通信模块、用户接口模块以及信息推荐程序。
在图1所示的信息推荐设备中，网络接口1004主要用于连接后台服务器，与所述后台服务器进行数据通信；用户接口1003主要用于连接用户设备；所述信息推荐设备通过处理器1001调用存储器1005中存储的信息推荐程序，并执行本申请实施例提供的信息推荐方法。
基于上述硬件结构，提出本申请信息推荐方法的实施例。
参照图2，图2为本申请信息推荐方法第一实施例的流程示意图，提出本申请信息推荐方法第一实施例。
在第一实施例中，所述信息推荐方法包括以下操作：
操作S10：获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息。
应理解的是，本实施例的执行主体是所述信息推荐设备，其中，所述信息推荐设备可为智能手机、个人电脑或服务器等电子设备，本实施例对此不加以限制。所述个人信息包括：姓名、年龄、身份证号码及性别等基础信息。所述待推荐用户通过终端输入所述预问诊信息，所述终端可以是所述待推荐用户的手机，也可以是医院的挂号设备，也可以是所述信息推荐设备。所述待推荐用户通过所述信息推荐设备录入的当前身体状况信息或想挂的科室等信息，可以通过所述信息推荐设备的显示界面展示的选项进行选择而生成所述预问诊信息，还可以是所述待推荐用户通过语音说出自己当前的身体状况信息，具体包括：是否发烧、咳嗽、流鼻涕、腰痛、头痛或眼睛疼等身体状况信息，以及想挂内科、皮肤科或眼科等意向就诊科室信息。
操作S20：从医院信息系统中查找与所述个人信息对应的患者医疗数据。
可理解的是，通常在医院进行过治疗的病人，所述医院信息系统均会记录病人每次就诊的信息。所述患者医疗数据为所述待推荐用户历史就医的记录信息，包括历史就诊科室、所开的处方及基本就医情况等信息。所述医院信息系统中记录着各用户的个人信息及对应的患者医疗数据，则可从医院信息系统中查找与所述个人信息对应的患者医疗数据。
操作S30：对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词。
需要说明的是，通常同一个病人去医院就医，可能看的是同一种病，复发或者未痊愈，再次来医院进行复诊，则可对所述患者医疗数据按照就诊科室进行分类，所述就诊科室包括皮肤科、内科和外科等，并对各历史就诊科室能够治疗的病症等信息进行关键词提取，可获取各历史就诊科室对应的治疗病症信息进行分词处理，获得所述治疗病症信息的所有词语，计算各词语的词频-逆文档频率（Term Frequency-Inverse Document Frequency，TF-TDF）值，TF-TDF值越大，表明该词越重要，可将所有词语按照TF-TDF值从大到小进行排序，获取排在前面的预设数量的词作为就诊关键词。所述预设数量可以根据经验设置，比如为3。
操作S40：对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配。
应理解的是，所述预问诊信息包括所述待推荐用户的当前身体状况或想挂的就诊科室等信息，对所述预问诊信息进行分词处理，可通过基于词典的分词算法，也就是字符串匹配，将待匹配的字符串基于一定的算法策略，和一个足够大的词典进行字符串匹配，如果匹配命中，则可以分词。根据不同的匹配策略，又分为正向最大匹配法，逆向最大匹配法，双向匹配分词，全切分路径选择等，从而获得所述预问诊信息的所有词语。
在具体实现中，将所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词表示成向量形式，计算向量形式的所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词之间的余弦距离作为相似度，根据所述相似度判断所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词是否匹配，可通过设置预设第一相似度阈值，比如80%，在所述相似度超过所述预设第一相似度阈值时，将超过所述预设第一相似度阈值的所述相似度对应的历史就诊科室认定为匹配成功的科室。还可通过选取相似度最高的历史就诊科室作为匹配成功的科室进行推荐。
操作S50：确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
需要说明的是，匹配成功的历史就诊科室即为所述待推荐用户需要挂号的科室，则将其作为所述第一目标科室进行推荐，可以是将所述第一目标科室在所述信息推荐设备的显示界面进行显示，所述待推荐用户对所述第一目标科室进行确认，即可完成挂号操作。也可以是语音播放所述第一目标科室，还可以是将所述第一目标科室推荐至所述待推荐用户的用户终端，比如智能手机或智能手表等，则所述待推荐用户可根据所述第一目标科室进行后续挂号操作。
本实施例中，通过从医院信息系统中查找与待推荐用户的个人信息对应的患者医疗数据，并获取预问诊信息，通过将预问诊信息与患者医疗数据进行匹配，即将当前患者身体状况信息结合用户历史就诊的患者医疗数据，更准确地为用户推荐就诊科室，无需用户对挂号系统中默认目录的多层级选项进行逐层选择，直接便捷地为用户推荐目标科室，提高挂号效率，提升用户体验。
参照图3，图3为本申请信息推荐方法第二实施例的流程示意图，基于上述图2所示的第一实施例，提出本申请信息推荐方法的第二实施例。
在第二实施例中，所述操作S40之后，还包括：
操作S401：确定匹配失败后，从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词。
应理解的是，若所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词匹配失败，说明所述待推荐用户不是进行复诊，需要挂新的科室，则从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室的科室基本信息，对所述其他科室的科室基本信息进行分词处理，获得各所述其他科室的科室基本信息的所有词语，计算各词语的TF-TDF值，所述TF-TDF值越大，表明该词越重要，可将所有词语按照所述TF-TDF值从大到小进行排序，获取排在前面的预设数量的词作为就诊关键词，所述预设数量可以根据经验设置，比如为3，从而获得各其他科室对应的其他科室关键词。
操作S402：将所述预问诊信息的所有词语与各其他科室对应的其他科室关键词进行匹配。
可理解的是，将所述预问诊信息的所有词语与其他科室关键词表示成向量形式，计算向量形式的所述预问诊信息的所有词语与其他科室关键词之间的余弦距离作为相似度，根据所述相似度判断所述预问诊信息的所有词语与其他科室关键词是否匹配，可通过设置预设第二相似度阈值，比如80%，在所述相似度超过所述预设第二相似度阈值时，将超过所述预设第二相似度阈值的所述相似度对应的其他科室认定为匹配成功的科室。还可通过选取相似度最高的其他科室作为匹配成功的科室进行推荐。
操作S403：确定匹配成功后，将匹配成功的其他科室作为第二目标科室，将所述第二目标科室进行推荐。
需要说明的是，匹配成功的其他科室即为所述待推荐用户需要挂号的科室，则将其作为所述第二目标科室进行推荐，可以是将所述第二目标科室在所述信息推荐设备的显示界面进行显示，所述待推荐用户对所述第二目标科室进行确认，即可完成挂号操作。也可以是语音播放所述第二目标科室，还可以是将所述第二目标科室推荐至所述待推荐用户的用户终端，比如智能手机或智能手表等，则所述待推荐用户可根据所述第二目标科室进行后续挂号操作。
操作S404：确定匹配失败后，将所述预问诊信息发送至目标终端，以使医护人员通过所述目标终端基于所述预问诊信息推荐就诊科室。
在具体实现中，可能存在所述预问诊信息录入不准确的情况，导致未能匹配出合适的科室进行推荐，则可将所述预问诊信息发送至目标终端，所述目标终端可以是所述医护人员的个人计算机或者智能手机等，以使医护人员通过所述目标终端查看所述预问诊信息，根据所述预问诊信息为所述待推荐用户推荐就诊科室，提高待推荐用户的挂号效率。
在本实施例中，在未能从历史就诊科室找到合适的科室进行推荐时，通过获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词，根据所述预问诊信息和所述其他科室关键词为待推荐用户推荐合适的科室，提高待推荐用户的挂号效率。
参照图4，图4为本申请信息推荐方法第三实施例的流程示意图，基于上述第一实施例或第二实施例，提出本申请信息推荐方法的第三实施例。本实施例基于第一实施例进行说明。
在第三实施例中，所述操作S10之后，还包括：
操作S101：获取所述待推荐用户的当前照片。
应理解的是，所述待推荐用户通常为去医院看病的病人，在大楼门口、挂号处或者各个诊室门口安装高清摄像头，实时抓取患者的面部信息和表情信息，可以是在所述待推荐用户进入医院时，通过摄像头或者其他拍摄设备拍摄所述待推荐用户的所述当前照片。
操作S102：对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征。
需要说明的是，可对所述当前照片进行预处理，具体为人脸检测及配准、人脸切割和图像归一化。可采用基于候选区域的深度学习目标检测算法（Fast Region-based Convolutional Neural Networks，Fast-RCNN）来检测人脸，对所述当前照片进行人脸切割，获得所述待推荐用户的人脸图片。对所述人脸图片进行图像归一化，几何归一化的目的主要是将表情子图像变换为统一的尺寸，有利于表情特征的提取。
进一步地，所述操作S102，包括：
对所述当前照片进行几何归一化，获得几何归一化图片；
对所述几何归一化图片进行灰度归一化，获得灰度归一化图片；
通过主成分分析算法对所述灰度归一化图片进行特征提取，获得所述当前照片对应的待识别图片特征。
应理解的是，首先，对所述当前照片标定特征点，这里用[x,y] = ginput(3)函数来标定两眼和鼻子三个特征点，主要是用鼠标动手标定，获取三个特征点的坐标值。再根据左右两眼的坐标值旋转图像，以保证人脸方向的一致性。设两眼之间的距离为d，其中点为O。根据面部特征点和几何模型确定矩形特征区域，以O为基准，左右各剪切d，垂直方向各取0.5d和1.5d的矩形区域进行裁剪。对表情子区域图像进行尺度变换为统一的尺寸，更有利于表情特征的提取。把截取的图像统一规格为90*100的图像，实现图像的几何归一化，获得所述几何归一化图片。
可理解的是，还可对所述几何归一化图片进行灰度归一化，灰度归一化主要是增加图像的亮度，使图像的细节更加清楚，以减弱光线和光照强度的影响。可采用预设图像函数进行光照补偿，所述预设图像函数可以是image=255*imadjust(C/255,[0.3;1],[0;1])，获得所述灰度归一化图片。
需要说明的是，通过主成分分析(Principal Component Analysis，PCA) 算法对所述灰度归一化图片进行特征提取，主成分，就是线性系数，即投影方向，将坐标轴中心移到数据的中心，然后旋转坐标轴，使得数据在C1轴上的上的方差最大，即全部n个数据个体在该方向上的投影最为分散，则更多的信息被保留下来，C1成为第一主成分，C2第二主成分，找一个C2，使得C2与C1的协方差（相关系数）为0，以免与C1信息重叠，并且使数据在该方向的方差尽量最大，以此类推，找到第三主成分，第四主成分......第p个主成分。p个随机变量就有p个主成分。通过协方差对特征值及特征向量进行分析，特征向量即为特征脸，以获得所述待识别图片特征。基于动态图片，还可采用几何法或深度学习法。
操作S103：根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情。
可理解的是，通过微表情检测预测患者的心理状态，例如患者是高兴、难受、发怒或者悲伤等。医院工作人员根据患者心理状态，以不同的态度对不同的病人，减少医患纠纷。例如：当患者处于发怒的心理状态时，医护人员尽量用温和的语气，避免激怒患者，引起伤医事件发生；当患者处于悲伤的心理状态时，医护人员尽量不要轻视病人，多说一些鼓励的话语，避免患者更加悲伤，甚至引发自杀事件。
进一步地，所述操作S103，包括：
通过递归神经网络模型对所述待识别图片特征进行学习，并通过随机森林模型对学习的特征进行分类，获得所述待推荐用户的当前表情。
应理解的是，微表情检测，采用深度学习通过多层网络结构，比如递归神经网络（Recurrent Neuron Networks，RNNs）进行多种非线性变换和表示，提取图片的高级抽象特征，在学习深度特征之后，最后一步是识别测试人脸的表情属于基本表情的哪一类。深度神经网络可以端到端地进行人脸表情识别。一种方式是在网络的末端加上损失层，来修正反向传播误差，每个样本的预测概率可以直接从网络中输出。另一种方式是利用深度神经网络作为提取特征的工具，然后再用传统的分类器，例如随机森林模型，对提取的特征进行分类，从而获得所述待推荐用户的当前表情。
操作S104：根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端。
需要说明的是，检测出所述待推荐用户的当前表情，所述当前表情包括高兴、难受、发怒、悲伤和不耐烦等。针对各种表情可预先设置对应的服务态度建议，比如，所述当前表情为发怒时，对应的服务态度建议为：用温和的语气，避免激怒患者。所述目标终端可以是医护人员的智能手机或者计算机，以使医护人员及时了解患者的心情，找到合适的方式进行沟通，以提高治疗的效率。
进一步地，在本实施例中，所述获取待推荐用户的个人信息，包括：
从公安系统库获取照片集，对所述照片集中各照片进行特征提取，获得各照片对应的照片特征；
将所述待识别图片特征与各照片特征进行匹配，确定匹配成功后，将匹配成功的照片对应的目标用户认定为所述待推荐用户；
从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息。
应理解的是，各照片对应的照片特征可以被用来将人脸对齐到平均人脸，这样在对齐之后所有图像中的人脸特征点的位置几乎是相同的。直观上来看，用对齐后的图像训练的人脸识别算法更加有效。对待识别图片特征进行面部特征点定位，获得待识别图片特征对应的待处理人脸特征点；将所述待处理人脸特征点与预设正脸特征点进行比较，获得单应性矩阵；通过所述单应性矩阵对照片中的人脸进行变换，获得校准人脸图片；通过卷积神经网络模型对所述校准人脸图片和安系统库中的各照片特征进行比对，获得待识别图片特征与各照片特征之间的人脸相似度。若所述人脸相似度超过预设人脸相似度阈值，则认为匹配成功，将匹配成功的公安系统库中的照片对应的用户作为所述目标用户。所述预设人脸相似度阈值可根据经验值进行设置，比如80%。
进一步地，在本实施例中，所述从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息之后，所述信息推荐方法还包括：
根据所述待推荐用户的个人信息判断所述待推荐用户是否为医患风险对象，获得判断结果；
将所述判断结果发送至所述目标终端。
可理解的是，通过人脸识别到公安系统库获取所述待推荐用户的个人信息，如果所述待推荐用户有过犯罪记录，则判定所述待推荐用户是医患风险对象，则将所述判断结果发送至所述目标终端，所述目标终端可以是医护人员的智能手机或者计算机等，以通知保安人员和医务人员重点关注，防止伤医事件发生。当微表情检测到患者处于伤医前期状态时，主动通知保安和公安机关，让他们前来防护制止，并通知医务人员做好防护措施，保障自身安全和医院财产安全。并且可以和医疗系统医闹数据库比对，针对医闹人员，保安和医院人员重点防范，做好防护措施，主动收集相关证据，防止医闹事件的发生。针对医闹事件，可以将收集的数据提交给管理机构和司法机关，有理有据的打击医闹人员。
本实施例中，通过获取所述待推荐用户的当前照片，对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征，根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情，根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端，以使医护人员及时了解患者的心情，找到合适的方式进行沟通，以提高治疗的效率。
此外，本申请实施例还提出一种计算机可读存储介质，所述计算机可读存储介质上存储有信息推荐程序，所述信息推荐程序被处理器执行时实现如上文所述的信息推荐方法的操作。所述计算机可读存储介质可以是非易失性，也可以是易失性，所述计算机可读存储介质可以简称为“存储介质”。
此外，参照图5，本申请实施例还提出一种信息推荐装置，所述信息推荐装置包括：
获取模块10，配置为获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息。
应理解的是，所述个人信息包括：姓名、年龄、身份证号码及性别等基础信息。所述待推荐用户通过终端输入所述预问诊信息，所述终端可以是所述待推荐用户的手机，也可以是医院的挂号设备，也可以是所述信息推荐设备。所述待推荐用户通过所述信息推荐设备录入的当前身体状况信息或想挂的科室等信息，可以通过所述信息推荐设备的显示界面展示的选项进行选择而生成所述预问诊信息，还可以是所述待推荐用户通过语音说出自己当前的身体状况信息，具体包括：是否发烧、咳嗽、流鼻涕、腰痛、头痛或眼睛疼等身体状况信息，以及想挂内科、皮肤科或眼科等意向就诊科室信息。
查找模块20，配置为从医院信息系统中查找与所述个人信息对应的患者医疗数据。
可理解的是，通常在医院进行过治疗的病人，所述医院信息系统均会记录病人每次就诊的信息。所述患者医疗数据为所述待推荐用户历史就医的记录信息，包括历史就诊科室、所开的处方及基本就医情况等信息。所述医院信息系统中记录着各用户的个人信息及对应的患者医疗数据，则可从医院信息系统中查找与所述个人信息对应的患者医疗数据。
分类模块30，配置为对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词。
需要说明的是，通常同一个病人去医院就医，可能看的是同一种病，复发或者未痊愈，再次来医院进行复诊，则可对所述患者医疗数据按照就诊科室进行分类，所述就诊科室包括皮肤科、内科和外科等，并对各历史就诊科室能够治疗的病症等信息进行关键词提取，可获取各历史就诊科室对应的治疗病症信息进行分词处理，获得所述治疗病症信息的所有词语，计算各词语的词频-逆文档频率（Term Frequency-Inverse Document Frequency，TF-TDF）值，TF-TDF值越大，表明该词越重要，可将所有词语按照TF-TDF值从大到小进行排序，获取排在前面的预设数量的词作为就诊关键词。所述预设数量可以根据经验设置，比如为3。
匹配模块40，配置为对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配。
应理解的是，所述预问诊信息包括所述待推荐用户的当前身体状况或想挂的就诊科室等信息，对所述预问诊信息进行分词处理，可通过基于词典的分词算法，也就是字符串匹配，将待匹配的字符串基于一定的算法策略，和一个足够大的词典进行字符串匹配，如果匹配命中，则可以分词。根据不同的匹配策略，又分为正向最大匹配法，逆向最大匹配法，双向匹配分词，全切分路径选择等，从而获得所述预问诊信息的所有词语。
在具体实现中，将所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词表示成向量形式，计算向量形式的所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词之间的余弦距离作为相似度，根据所述相似度判断所述预问诊信息的所有词语与各历史就诊科室对应的就诊关键词是否匹配，可通过设置预设第一相似度阈值，比如80%，在所述相似度超过所述预设第一相似度阈值时，将超过所述预设第一相似度阈值的所述相似度对应的历史就诊科室认定为匹配成功的科室。还可通过选取相似度最高的历史就诊科室作为匹配成功的科室进行推荐。
推荐模块50，配置为确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
需要说明的是，匹配成功的历史就诊科室即为所述待推荐用户需要挂号的科室，则将其作为所述第一目标科室进行推荐，可以是将所述第一目标科室在所述信息推荐设备的显示界面进行显示，所述待推荐用户对所述第一目标科室进行确认，即可完成挂号操作。也可以是语音播放所述第一目标科室，还可以是将所述第一目标科室推荐至所述待推荐用户的用户终端，比如智能手机或智能手表等，则所述待推荐用户可根据所述第一目标科室进行后续挂号操作。
本实施例中，通过从医院信息系统中查找与待推荐用户的个人信息对应的患者医疗数据，并获取预问诊信息，通过将预问诊信息与患者医疗数据进行匹配，即将当前患者身体状况信息结合用户历史就诊的患者医疗数据，更准确地为用户推荐就诊科室，无需用户对挂号系统中默认目录的多层级选项进行逐层选择，直接便捷地为用户推荐目标科室，提高挂号效率，提升用户体验。
在一实施例中，所述信息推荐装置还包括：
所述获取模块10，还配置为确定匹配失败后，从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词；
所述匹配模块40，还配置为将所述预问诊信息的所有词语与各其他科室对应的其他科室关键词进行匹配；
所述推荐模块50，还配置为确定匹配成功后，将匹配成功的其他科室作为第二目标科室，将所述第二目标科室进行推荐；
发送模块，配置为确定匹配失败后，将所述预问诊信息发送至目标终端，以使医护人员通过所述目标终端基于所述预问诊信息推荐就诊科室。
在一实施例中，所述信息推荐装置还包括：
所述获取模块10，还配置为获取所述待推荐用户的当前照片；
特征提取模块，配置为对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征；
微表情检测模块，配置为根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情；
所述查找模块20，还配置为根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端。
在一实施例中，所述特征提取模块，配置为对所述当前照片进行几何归一化，获得几何归一化图片；对所述几何归一化图片进行灰度归一化，获得灰度归一化图片；通过主成分分析算法对所述灰度归一化图片进行特征提取，获得所述当前照片对应的待识别图片特征。
在一实施例中，所述分类模块30，还配置为通过递归神经网络模型对所述待识别图片特征进行学习，并通过随机森林模型对学习的特征进行分类，获得所述待推荐用户的当前表情。
在一实施例中，所述获取模块10，还配置为从公安系统库获取照片集，对所述照片集中各照片进行特征提取，获得各照片对应的照片特征；将所述待识别图片特征与各照片特征进行匹配，确定匹配成功后，将匹配成功的照片对应的目标用户认定为所述待推荐用户；从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息。
在一实施例中，所述信息推荐装置还包括：
判断模块，配置为根据所述待推荐用户的个人信息判断所述待推荐用户是否为医患风险对象，获得判断结果；
所述发送模块，还配置为将所述判断结果发送至所述目标终端。
本申请所述信息推荐装置的其他实施例或具体实现方式可参照上述各方法实施例，此处不再赘述。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者系统不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者系统所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者系统中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。词语第一、第二、以及第三等的使用不表示任何顺序，可将这些词语解释为标识。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如只读存储器镜像（Read Only Memory image，ROM）/随机存取存储器（Random Access Memory，RAM）、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (22)

  1. 一种信息推荐方法，所述信息推荐方法包括以下操作：
    获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
    从医院信息系统中查找与所述个人信息对应的患者医疗数据；
    对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
    对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
    确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
  2. 如权利要求1所述的信息推荐方法，其中，在所述“对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配”之后，所述信息推荐方法还包括：
    确定匹配失败后，从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词；
    将所述预问诊信息的所有词语与各其他科室对应的其他科室关键词进行匹配；
    确定匹配成功后，将匹配成功的其他科室作为第二目标科室，将所述第二目标科室进行推荐；
    确定匹配失败后，将所述预问诊信息发送至目标终端，以使医护人员通过所述目标终端基于所述预问诊信息推荐就诊科室。
  3. 如权利要求1所述的信息推荐方法，其中，所述“获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息”之后，所述信息推荐方法还包括：
    获取所述待推荐用户的当前照片；
    对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征；
    根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情；
    根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端。
  4. 如权利要求3所述的信息推荐方法，其中，所述“对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征”，包括：
    对所述当前照片进行几何归一化，获得几何归一化图片；
    对所述几何归一化图片进行灰度归一化，获得灰度归一化图片；
    通过主成分分析算法对所述灰度归一化图片进行特征提取，获得所述当前照片对应的待识别图片特征。
  5. 如权利要求3所述的信息推荐方法，其中，所述“根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情”，包括：
    通过递归神经网络模型对所述待识别图片特征进行学习，并通过随机森林模型对学习的特征进行分类，获得所述待推荐用户的当前表情。
  6. 如权利要求1-5中任一项所述的信息推荐方法，其中，所述“获取待推荐用户的个人信息”，包括：
    从公安系统库获取照片集，对所述照片集中各照片进行特征提取，获得各照片对应的照片特征；
    将所述待识别图片特征与各照片特征进行匹配，确定匹配成功后，将匹配成功的照片对应的目标用户认定为所述待推荐用户；
    从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息。
  7. 如权利要求6所述的信息推荐方法，其中，所述“从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息”之后，所述信息推荐方法还包括：
    根据所述待推荐用户的个人信息判断所述待推荐用户是否为医患风险对象，获得判断结果；
    将所述判断结果发送至所述目标终端。
  8. 一种信息推荐设备，所述信息推荐设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的信息推荐程序，所述信息推荐程序被所述处理器执行时实现如下操作：
    获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
    从医院信息系统中查找与所述个人信息对应的患者医疗数据；
    对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
    对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
    确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
  9. 如权利要求8所述的信息推荐设备，其中，所述信息推荐程序被所述处理器执行时，在所述“对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配”之后，还实现如下操作：
    确定匹配失败后，从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词；
    将所述预问诊信息的所有词语与各其他科室对应的其他科室关键词进行匹配；
    确定匹配成功后，将匹配成功的其他科室作为第二目标科室，将所述第二目标科室进行推荐；
    确定匹配失败后，将所述预问诊信息发送至目标终端，以使医护人员通过所述目标终端基于所述预问诊信息推荐就诊科室。
  10. 如权利要求8所述的信息推荐设备，其中，所述信息推荐程序被所述处理器执行时，在所述“获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息”之后，还实现如下操作：
    获取所述待推荐用户的当前照片；
    对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征；
    根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情；
    根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端。
  11. 如权利要求10所述的信息推荐设备，其中，所述“对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征”，包括：
    对所述当前照片进行几何归一化，获得几何归一化图片；
    对所述几何归一化图片进行灰度归一化，获得灰度归一化图片；
    通过主成分分析算法对所述灰度归一化图片进行特征提取，获得所述当前照片对应的待识别图片特征。
  12. 如权利要求10所述的信息推荐设备，其中，所述“根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情”，包括：
    通过递归神经网络模型对所述待识别图片特征进行学习，并通过随机森林模型对学习的特征进行分类，获得所述待推荐用户的当前表情。
  13. 如权利要求8-12中任一项所述的信息推荐设备，其中，所述“获取待推荐用户的个人信息”，包括：
    从公安系统库获取照片集，对所述照片集中各照片进行特征提取，获得各照片对应的照片特征；
    将所述待识别图片特征与各照片特征进行匹配，确定匹配成功后，将匹配成功的照片对应的目标用户认定为所述待推荐用户；
    从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息。
  14. 如权利要求13所述的信息推荐设备，其中，所述信息推荐程序被所述处理器执行时，所述“从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息”之后，还实现如下操作：
    根据所述待推荐用户的个人信息判断所述待推荐用户是否为医患风险对象，获得判断结果；
    将所述判断结果发送至所述目标终端。
  15. 一种计算机可读存储介质，所述计算机可读存储介质上存储有信息推荐程序，所述信息推荐程序被处理器执行时实现如下操作：
    获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
    从医院信息系统中查找与所述个人信息对应的患者医疗数据；
    对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
    对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
    确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
  16. 如权利要求15所述的计算机可读存储介质，其中，所述信息推荐程序被所述处理器执行时，在所述“对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配”之后，还实现如下操作：
    确定匹配失败后，从所述医院信息系统中获取除了所述历史就诊科室之外的其他科室和对应的其他科室关键词；
    将所述预问诊信息的所有词语与各其他科室对应的其他科室关键词进行匹配；
    确定匹配成功后，将匹配成功的其他科室作为第二目标科室，将所述第二目标科室进行推荐；
    确定匹配失败后，将所述预问诊信息发送至目标终端，以使医护人员通过所述目标终端基于所述预问诊信息推荐就诊科室。
  17. 如权利要求15所述的计算机可读存储介质，其中，所述信息推荐程序被所述处理器执行时，在所述“获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息”之后，还实现如下操作：
    获取所述待推荐用户的当前照片；
    对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征；
    根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情；
    根据所述当前表情查找对应的服务态度建议，将所述服务态度建议发送至目标终端。
  18. 如权利要求17所述的计算机可读存储介质，其中，所述“对所述当前照片进行特征提取，获得所述当前照片对应的待识别图片特征”，包括：
    对所述当前照片进行几何归一化，获得几何归一化图片；
    对所述几何归一化图片进行灰度归一化，获得灰度归一化图片；
    通过主成分分析算法对所述灰度归一化图片进行特征提取，获得所述当前照片对应的待识别图片特征。
  19. 如权利要求17所述的计算机可读存储介质，其中，所述“根据所述待识别图片特征进行微表情检测，获得所述待推荐用户的当前表情”，包括：
    通过递归神经网络模型对所述待识别图片特征进行学习，并通过随机森林模型对学习的特征进行分类，获得所述待推荐用户的当前表情。
  20. 如权利要求15-19中任一项所述的计算机可读存储介质，其中，所述“获取待推荐用户的个人信息”，包括：
    从公安系统库获取照片集，对所述照片集中各照片进行特征提取，获得各照片对应的照片特征；
    将所述待识别图片特征与各照片特征进行匹配，确定匹配成功后，将匹配成功的照片对应的目标用户认定为所述待推荐用户；
    从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息。
  21. 如权利要求20所述的计算机可读存储介质，其中，所述信息推荐程序被所述处理器执行时，在所述“从公安系统库中获取所述目标用户的个人信息作为所述待推荐用户的个人信息”之后，还实现如下操作：
    根据所述待推荐用户的个人信息判断所述待推荐用户是否为医患风险对象，获得判断结果；
    将所述判断结果发送至所述目标终端。
  22. 一种信息推荐装置，所述信息推荐装置包括：
    获取模块，配置为获取待推荐用户的个人信息，并获取所述待推荐用户的预问诊信息；
    查找模块，配置为从医院信息系统中查找与所述个人信息对应的患者医疗数据；
    分类模块，配置为对所述患者医疗数据按照就诊科室进行分类，获得所述待推荐用户的历史就诊科室和各历史就诊科室对应的就诊关键词；
    匹配模块，配置为对所述预问诊信息进行分词处理，并将对所述预问诊信息进行分词处理得到的所有词语与各历史就诊科室对应的就诊关键词分别进行匹配；
    推荐模块，配置为确定匹配成功后，将匹配成功的历史就诊科室作为第一目标科室，并将所述第一目标科室进行推荐。
PCT/CN2020/106345 2020-01-09 2020-07-31 信息推荐方法、设备、计算机可读存储介质及装置 WO2021139146A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010024389.9A CN111241265A (zh) 2020-01-09 2020-01-09 信息推荐方法、设备、存储介质及装置
CN202010024389.9 2020-01-09

Publications (1)

Publication Number Publication Date
WO2021139146A1 true WO2021139146A1 (zh) 2021-07-15

Family

ID=70866104

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/106345 WO2021139146A1 (zh) 2020-01-09 2020-07-31 信息推荐方法、设备、计算机可读存储介质及装置

Country Status (2)

Country Link
CN (1) CN111241265A (zh)
WO (1) WO2021139146A1 (zh)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113488159A (zh) * 2021-08-11 2021-10-08 中国医学科学院阜外医院 一种基于神经网络的医疗科室推荐方法及装置
CN113744828A (zh) * 2021-08-31 2021-12-03 平安国际智慧城市科技股份有限公司 一种病历推荐方法、装置、设备及存储介质
CN113823393A (zh) * 2021-11-22 2021-12-21 中哲国际工程设计有限公司 基于bim技术的医院就医导航系统及方法
CN113840006A (zh) * 2021-09-27 2021-12-24 平安国际智慧城市科技股份有限公司 问诊接口的管理方法、装置、电子设备以及存储介质
CN115271851A (zh) * 2022-07-04 2022-11-01 天翼爱音乐文化科技有限公司 一种视频彩铃推荐方法、系统、电子设备及存储介质
CN115329156A (zh) * 2022-10-14 2022-11-11 北京云成金融信息服务有限公司 基于历史数据的数据治理方法及系统
CN116501990A (zh) * 2023-04-11 2023-07-28 北京师范大学-香港浸会大学联合国际学院 基于门诊大数据的医院专科影响力评估方法及装置
CN116564538A (zh) * 2023-07-05 2023-08-08 肇庆市高要区人民医院 一种基于大数据的医院就医信息实时查询方法及系统
CN116895358A (zh) * 2023-09-11 2023-10-17 江苏泰德医药有限公司 一种基于云平台医疗资源智能管理系统及方法
CN116955832A (zh) * 2023-09-19 2023-10-27 中科数创(北京)数字传媒有限公司 一种基于大数据的科普知识精准推送方法及系统
CN116959686A (zh) * 2023-07-27 2023-10-27 常州云燕医疗科技有限公司 一种基于数字一体化的医疗信息管理系统及方法
CN117112628A (zh) * 2023-09-08 2023-11-24 廊坊丛林科技有限公司 一种物流数据的更新方法及系统
CN117636265A (zh) * 2024-01-25 2024-03-01 南昌大学第一附属医院 一种用于医疗场景中的患者定位方法及系统
CN117709493A (zh) * 2023-12-26 2024-03-15 江苏道朴网络科技有限公司 一种基于历史病理的智能医疗问诊系统

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111241265A (zh) * 2020-01-09 2020-06-05 平安国际智慧城市科技股份有限公司 信息推荐方法、设备、存储介质及装置
TWI795651B (zh) * 2020-06-30 2023-03-11 廖珮宏 引導式智慧門診掛號輔助系統及方法
CN112185498B (zh) * 2020-09-24 2022-10-14 毛真真 一种中医内科患者数据信息处理方法、系统、装置
CN112364065B (zh) * 2020-10-27 2022-04-15 刘锋 一种获取大数据中转、反馈的方法及系统
CN112365981A (zh) * 2020-11-26 2021-02-12 中国联合网络通信集团有限公司 智慧医疗信息处理方法和装置
CN112786176A (zh) * 2021-02-22 2021-05-11 北京融威众邦电子技术有限公司 智能自助就诊方法、装置、计算机设备
CN113130052A (zh) * 2021-03-09 2021-07-16 深圳星医科技有限公司 医生推荐方法、医生推荐装置、终端设备及存储介质
CN113782163A (zh) * 2021-03-29 2021-12-10 北京京东拓先科技有限公司 信息推送方法、装置和计算机可读存储介质
CN113066567B (zh) * 2021-04-09 2022-09-09 武汉市同步远方信息技术开发有限公司 一种基于区块链的医疗就诊挂号系统
CN113160914A (zh) * 2021-04-26 2021-07-23 联仁健康医疗大数据科技股份有限公司 在线问诊方法、装置、电子设备及存储介质
CN113241134B (zh) * 2021-04-26 2023-07-04 哈尔滨工业大学（深圳） 在线问诊医生分配方法和系统、存储介质
CN113496774A (zh) * 2021-07-22 2021-10-12 河北北方学院 一种数字化的医疗信息管理方法及相关装置
CN113707301A (zh) * 2021-08-30 2021-11-26 康键信息技术（深圳）有限公司 基于人工智能的远程问诊方法、装置、设备及介质
CN113707335A (zh) * 2021-09-06 2021-11-26 挂号网（杭州）科技有限公司 确定目标接诊用户的方法、装置、电子设备和存储介质
CN114360137A (zh) * 2021-11-30 2022-04-15 浙江朱道模块集成有限公司 一种用于医院的信息采集自动引导系统
CN114512241B (zh) * 2021-12-27 2024-05-03 中国人民解放军总医院第一医学中心 一种基于频次分析的食管静脉瘤信息智能搜寻方法及系统
CN115497615B (zh) * 2022-10-24 2023-09-01 北京亿家老小科技有限公司 一种远程医疗方法及系统
CN116913493A (zh) * 2023-07-24 2023-10-20 北京同仁堂互联网医院管理有限公司 一种医患的匹配方法、装置、设备及可读介质
CN116682544B (zh) * 2023-08-03 2023-10-20 南通大学附属医院 一种人工智能医疗自动化测试集成系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170018020A1 (en) * 2015-07-15 2017-01-19 Fuji Xerox Co., Ltd. Information processing apparatus and method and non-transitory computer readable medium
CN108231174A (zh) * 2017-12-11 2018-06-29 浪潮软件集团有限公司 一种确定科室的方法、装置及系统
CN108922608A (zh) * 2018-06-13 2018-11-30 平安医疗科技有限公司 智能导诊方法、装置、计算机设备和存储介质
CN109599187A (zh) * 2018-10-31 2019-04-09 北京春雨天下软件有限公司 一种在线问诊的分诊方法、服务器、终端、设备及介质
CN111241265A (zh) * 2020-01-09 2020-06-05 平安国际智慧城市科技股份有限公司 信息推荐方法、设备、存储介质及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170018020A1 (en) * 2015-07-15 2017-01-19 Fuji Xerox Co., Ltd. Information processing apparatus and method and non-transitory computer readable medium
CN108231174A (zh) * 2017-12-11 2018-06-29 浪潮软件集团有限公司 一种确定科室的方法、装置及系统
CN108922608A (zh) * 2018-06-13 2018-11-30 平安医疗科技有限公司 智能导诊方法、装置、计算机设备和存储介质
CN109599187A (zh) * 2018-10-31 2019-04-09 北京春雨天下软件有限公司 一种在线问诊的分诊方法、服务器、终端、设备及介质
CN111241265A (zh) * 2020-01-09 2020-06-05 平安国际智慧城市科技股份有限公司 信息推荐方法、设备、存储介质及装置

Cited By (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113488159A (zh) * 2021-08-11 2021-10-08 中国医学科学院阜外医院 一种基于神经网络的医疗科室推荐方法及装置
CN113744828A (zh) * 2021-08-31 2021-12-03 平安国际智慧城市科技股份有限公司 一种病历推荐方法、装置、设备及存储介质
CN113744828B (zh) * 2021-08-31 2023-06-02 深圳平安智慧医健科技有限公司 一种病历推荐方法、装置、设备及存储介质
CN113840006A (zh) * 2021-09-27 2021-12-24 平安国际智慧城市科技股份有限公司 问诊接口的管理方法、装置、电子设备以及存储介质
CN113840006B (zh) * 2021-09-27 2023-07-11 深圳平安智慧医健科技有限公司 问诊接口的管理方法、装置、电子设备以及存储介质
CN113823393A (zh) * 2021-11-22 2021-12-21 中哲国际工程设计有限公司 基于bim技术的医院就医导航系统及方法
CN113823393B (zh) * 2021-11-22 2022-04-22 中哲国际工程设计有限公司 基于bim技术的医院就医导航系统及方法
CN115271851A (zh) * 2022-07-04 2022-11-01 天翼爱音乐文化科技有限公司 一种视频彩铃推荐方法、系统、电子设备及存储介质
CN115271851B (zh) * 2022-07-04 2023-10-10 天翼爱音乐文化科技有限公司 一种视频彩铃推荐方法、系统、电子设备及存储介质
CN115329156A (zh) * 2022-10-14 2022-11-11 北京云成金融信息服务有限公司 基于历史数据的数据治理方法及系统
CN116501990B (zh) * 2023-04-11 2024-01-26 北京师范大学-香港浸会大学联合国际学院 基于门诊大数据的医院专科影响力评估方法及装置
CN116501990A (zh) * 2023-04-11 2023-07-28 北京师范大学-香港浸会大学联合国际学院 基于门诊大数据的医院专科影响力评估方法及装置
CN116564538A (zh) * 2023-07-05 2023-08-08 肇庆市高要区人民医院 一种基于大数据的医院就医信息实时查询方法及系统
CN116564538B (zh) * 2023-07-05 2023-12-19 肇庆市高要区人民医院 一种基于大数据的医院就医信息实时查询方法及系统
CN116959686A (zh) * 2023-07-27 2023-10-27 常州云燕医疗科技有限公司 一种基于数字一体化的医疗信息管理系统及方法
CN116959686B (zh) * 2023-07-27 2024-02-13 常州云燕医疗科技有限公司 一种基于数字一体化的医疗信息管理系统及方法
CN117112628A (zh) * 2023-09-08 2023-11-24 廊坊丛林科技有限公司 一种物流数据的更新方法及系统
CN116895358B (zh) * 2023-09-11 2023-11-28 江苏泰德医药有限公司 一种基于云平台医疗资源智能管理系统及方法
CN116895358A (zh) * 2023-09-11 2023-10-17 江苏泰德医药有限公司 一种基于云平台医疗资源智能管理系统及方法
CN116955832A (zh) * 2023-09-19 2023-10-27 中科数创(北京)数字传媒有限公司 一种基于大数据的科普知识精准推送方法及系统
CN116955832B (zh) * 2023-09-19 2023-11-28 中科数创(北京)数字传媒有限公司 一种基于大数据的科普知识精准推送方法及系统
CN117709493A (zh) * 2023-12-26 2024-03-15 江苏道朴网络科技有限公司 一种基于历史病理的智能医疗问诊系统
CN117636265A (zh) * 2024-01-25 2024-03-01 南昌大学第一附属医院 一种用于医疗场景中的患者定位方法及系统

Also Published As

Publication number Publication date
CN111241265A (zh) 2020-06-05

Similar Documents

Publication Publication Date Title
WO2021139146A1 (zh) 信息推荐方法、设备、计算机可读存储介质及装置
US11631175B2 (en) AI-based heat map generating system and methods for use therewith
US11282595B2 (en) Heat map generating system and methods for use therewith
US20200357118A1 (en) Medical scan viewing system with enhanced training and methods for use therewith
Mason et al. An investigation of biometric authentication in the healthcare environment
US11462318B2 (en) Method and system for computer-aided triage of stroke
JP2009513205A (ja) 特に診断画像に使用する画像処理システム
JP2014071610A (ja) データ処理装置、名寄せ処理方法及びコンピュータプログラム
CN110752027B (zh) 电子病历数据推送方法、装置、计算机设备和存储介质
Barakat et al. A machine learning approach on chest X-rays for pediatric pneumonia detection
Harisinghani et al. Classification of Alzheimer's using Deep-learning Methods on Webcam-based Gaze Data
CN116130088A (zh) 多模态面诊问诊方法、装置及相关设备
KR102594093B1 (ko) 딥러닝 기술을 활용한 피부과 시술 추천 시스템 및 방법
JP2004151820A (ja) 迷子検索・監視システム
Ampamya et al. Performance of an open source facial recognition system for unique patient matching in a resource-limited setting
Pabiania et al. Face recognition system for electronic medical record to access out-patient information
Adithama et al. Implementation of Face Recognition for Patient Identification Using the Transfer Learning Method
Shamova Face recognition in healthcare: general overview
KR102510599B1 (ko) 익명화된 의료정보에 대한 2차적 의학 소견의 생성 및 관리를 위한 클라우드 컴퓨팅 환경기반 네트워크 서비스 시스템 및 방법
KR102344848B1 (ko) 안면 인식을 이용한 환자 식별 장치 및 방법
Tsai et al. Iris recognition based on relative variation analysis with feature selection
Matey et al. Forensic Iris: A Review, 2022
US20240119175A1 (en) Machine learning for data anonymization
US20220208317A1 (en) Image content extraction method and image content extraction device
Sandamal et al. Emergency Patient Identification System

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20911685

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20911685

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 15.03.2023)

122 Ep: pct application non-entry in european phase

Ref document number: 20911685

Country of ref document: EP

Kind code of ref document: A1