Abnormal patient identification method, device, terminal device and storage medium

Abstract

This application is applicable to the field of artificial intelligence technology, and provides an abnormal patient recognition method, device, terminal device and storage medium. By performing vector representation on the name, type and cost of each charging item for each patient visit, a representation vector of the data for each patient visit is obtained; in chronological order, the representation vectors of the data for each patient visit of the patient are input into a long short-term memory network to obtain a representation vector of the data for multiple patient visits of the patient; based on the representation vectors of the multiple patient visit data of the patient, it is identified whether the patient is abnormal. By mining the multiple patient visit behaviors of the patient, abnormal patient recognition is carried out, which improves the recognition accuracy and enriches the abnormal patient recognition method library.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to a method, device, terminal device, and storage medium for identifying abnormal patients. Background Art

[0002] With the continuous deepening of the integration of artificial intelligence technology and the medical field, more and more medical-related tasks rely on advanced artificial intelligence technologies (including deep learning technology, machine learning technology, natural language processing technology, big data analysis technology, etc.) to be realized. Medical quality control refers to the identification, analysis, evaluation, and processing of existing and potential risks in medical activities, and the planned and organized reduction and elimination of the occurrence of risks, and the reduction of the adverse effects and economic losses caused by risk events. In the field of medical quality control, the identification of abnormal patients is one of the important tasks.

[0003] The existing methods for identifying abnormal patients mainly analyze the basic information of patients (such as age, gender, height, weight, etc.) and identify abnormal patients based on statistical values. However, this method is relatively basic and has limited ability to identify abnormal patients. Summary of the Invention

[0004] In view of this, the embodiments of this application provide a method, device, terminal device, and storage medium for identifying abnormal patients to solve the problem that the existing methods for identifying abnormal patients mainly analyze the basic information of patients and identify abnormal patients based on statistical values, and the ability to identify abnormal patients is limited.

[0005] The first aspect of the embodiments of this application provides a method for identifying abnormal patients, including:

[0006] Performing vector representation on the data of each patient visit to obtain a representation vector of the data of each patient visit, where the data of each patient visit includes the name, type, and cost of each charge item;

[0007] Inputting the representation vectors of the data of each patient visit into a long short-term memory network in chronological order to obtain a representation vector of the data of multiple patient visits;

[0008] Identifying whether the patient is abnormal based on the representation vectors of the multiple patient visit data.

[0009] The second aspect of the embodiments of this application provides an apparatus for identifying abnormal patients, including:

[0010] A first vector representation unit for performing vector representation on the data of each patient visit to obtain a representation vector of the data of each patient visit, where the data of each patient visit includes the name, type, and cost of each charge item;

[0011] A second vector representation unit, configured to input the representation vectors of the data of each visit of the patient into a long short-term memory network in chronological order, and obtain the representation vectors of the data of multiple visits of the patient;

[0012] An identification unit, configured to identify whether the patient is abnormal based on the representation vectors of the data of multiple visits of the patient.

[0013] A third aspect of the embodiments of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the terminal device. When the processor executes the computer program, the steps of the abnormal patient identification method provided in the first aspect are implemented.

[0014] A fourth aspect of the embodiments of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the abnormal patient identification method provided in the first aspect are implemented.

[0015] The abnormal patient identification method provided in the first aspect of the embodiments of the present application represents the name, type, and cost of each charge item for each visit of the patient to obtain the representation vectors of the data of each visit of the patient; input the representation vectors of the data of each visit of the patient into a long short-term memory network in chronological order to obtain the representation vectors of the data of multiple visits of the patient; identify whether the patient is abnormal based on the representation vectors of the data of multiple visits of the patient, and identify abnormal patients by mining the multiple visit behaviors of the patient, thereby improving the identification accuracy and enriching the abnormal patient identification method library.

[0016] It can be understood that the beneficial effects of the above second aspect to the fourth aspect can refer to the relevant descriptions in the above first aspect, and will not be elaborated here. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 is the first flowchart of the abnormal patient identification method provided by the embodiments of the present application;

[0019] Figure 2 is the second flowchart of the abnormal patient identification method provided by the embodiments of the present application;

[0020] Figure 3 It is a schematic structural diagram of a representation learning model for a patient's single visit behavior provided by an embodiment of the present application;

[0021] Figure 4 It is a schematic structural diagram of a long short-term memory network provided by an embodiment of the present application;

[0022] Figure 5 It is a schematic structural diagram of an abnormal patient identification device provided by an embodiment of the present application;

[0023] Figure 6 It is a schematic structural diagram of a terminal device provided by an embodiment of the present application. Detailed implementation manners

[0024] In the following description, for the purpose of illustration rather than limitation, specific details such as specific system structures and technologies are presented to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0025] It should be understood that when used in the specification and appended claims of the present application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0026] It should also be understood that the term "and / or" as used in the specification and appended claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0027] As used in the specification and appended claims of the present application, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if detecting [the described condition or event]" can be interpreted as meaning "once determined", "in response to determining", "once detecting [the described condition or event]", or "in response to detecting [the described condition or event]" according to the context.

[0028] In addition, in the description of the specification and appended claims of the present application, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0029] Reference to "one embodiment" or "some embodiments" etc. described in the specification of this application means that a specific feature, structure or characteristic described in connection with that embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments" etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized.

[0030] An embodiment of this application provides an abnormal patient recognition method, which can be executed by a processor of a terminal device when running a corresponding computer program. By mining the multiple medical visit behaviors of patients, abnormal patient recognition is performed, improving the recognition accuracy and enriching the abnormal patient recognition method library.

[0031] The abnormal patient recognition method provided by the embodiment of this application can be applied to the intelligent medical visit scenario, thereby promoting the construction of a smart city. For example, clinical medical quality control scenarios, medical insurance cost control scenarios, medical insurance risk control scenarios, etc.

[0032] In application, the terminal device can be a (cloud) server, a personal computer, a laptop computer and other computing devices capable of realizing data processing functions. For example, the server of a medical institution, which is used to run a Hospital Information System, and the abnormal patient recognition method is one of the functions that can be realized by this medical information management system.

[0033] In one embodiment, the abnormal patient recognition method provided by the embodiment of this application includes the following steps S101 to S103:

[0034] Step S101: Perform vector representation on the data of each medical visit of the patient to obtain a representation vector of the data of each medical visit of the patient.

[0035] In application, the user can input an instruction for obtaining the data of each medical visit of a specified patient through the human-computer interaction device of the terminal device, so that the terminal device obtains the data of each medical visit of the specified patient when responding to this instruction. For example, the user can input an instruction carrying the unique identity document (ID) of the patient to obtain the data of each medical visit of the corresponding patient; or, the user can also input an instruction for executing the abnormal patient recognition method through the human-computer interaction device of the terminal device, so that the terminal device automatically obtains the data of each medical visit of each patient by default when responding to this instruction.

[0036] In an application, the human-machine interaction device may include, but is not limited to, at least one of a keyboard, physical buttons, a touch sensor, a gesture recognition sensor, and a voice recognition unit, so that a user can control the terminal device through corresponding touch methods, gesture control methods, or voice control methods. The touch method for the physical buttons may specifically be pressing or toggling, and the touch method for the touch sensor may specifically be pressing or touching, etc. The gestures for controlling the terminal device can be custom-set by the user according to actual needs or adopt the default settings at the time of factory shipment. The voice recognition unit may include a microphone and a voice recognition chip, and the voice for controlling the terminal device can be custom-set by the user in advance through the human-machine interaction device of the terminal device or adopt the default settings at the time of factory shipment. The human-machine interaction device may be integrally provided in the terminal device. For example, the touch sensor may be integrally provided with the display of the terminal device as a touch display. The human-machine interaction device may also be an external device of the terminal device and be communicatively connected to the terminal device. For example, a keyboard and a microphone, etc., may be communicatively connected to the terminal device through the communication interface of the terminal device.

[0037] In an application, a patient's one-time medical visit behavior will involve multiple medical activities, and these medical activities can be reflected by specific charging items. Therefore, the abnormal patient identification method provided in the embodiments of the present application mines the patient's medical visit behavior based on the charging items generated in each medical visit behavior of the patient. A charging item usually has three attributes, namely name, type, and cost.

[0038] In an application, an association relationship can be established among various types of data (i.e., patient ID, and the name, type, and cost of the charging item) in the data of each medical visit of each patient. This association relationship may specifically be a mapping relationship and may exist in the form of an association relationship table. The association relationship table may specifically be a Look-Up-Table (LUT), or may exist in a form where corresponding search results can be found and output through other input data, so as to facilitate querying and calling. By establishing the association relationship in advance, when it is necessary to query and call the data of each medical visit of each patient, only one type of data in the data of each medical visit of the patient needs to be used to query and call other types of data associated therewith, thereby effectively saving the computing power resources and execution time of the terminal device.

[0039] As Figure 2 shown, in one embodiment, step S101 includes the following steps S201 to S205:

[0040] Step S201: Perform vector representation on the name of the charging item for each medical visit of the patient to obtain a representation vector of the name of each charging item for each medical visit of the patient, and the representation vectors of the same charging item name are the same;

[0041] Step S202: Represent the types of the charging items for each patient visit in vector form to obtain the representation vectors of the types of each charging item for each patient visit. The representation vectors of the same type of charging items are the same;

[0042] Step S203: Represent the expenses of the charging items for each patient visit in vector form to obtain the representation vectors of the expenses of the charging items for each patient visit. The representation vectors of the same expenses of the charging items are the same.

[0043] In application, in steps S201 to S203, the names, types, and expenses of the charging items for each patient visit are respectively represented in vector form to obtain the representation vectors of the names, types, and expenses of the charging items for each patient visit.

[0044] In application, based on the three attributes of the charging item name, type, and expense, three embedding layers can be respectively constructed to represent these three attributes in vector form, as follows:

[0045] Name embeddings, which are used to represent the names of the charging items in vector form. The representation vectors of the same names are the same;

[0046] Type embeddings, which are used to represent the types of the charging items (e.g., bed care fees, examination fees, test fees, etc.) in vector form. The representation vectors of the same types are the same;

[0047] Cost embeddings, which are used to represent the expenses of the charging items in vector form. The representation vectors of the same expenses are the same.

[0048] In application, the expenses of all the charging items for each patient visit are divided into intervals (e.g., divided into 100 intervals). Each interval corresponds to a spending representation vector. If the expense of a charging item falls within a certain interval, the vector representation of that interval is the first value; otherwise, it is the second value.

[0049] In application, the first value can specifically be 1, and the second value can specifically be 0. That is, a binary vector used as a machine language is used to represent whether the behavior of the patient is abnormal, which is conducive to the terminal device for identification. The user can also set the first value and the second value to other values through the human-computer interaction device of the terminal device according to actual needs.

[0050] In one embodiment, step S203 includes:

[0051] Represent the vector of the intervals into which the cost of each charge item for each patient visit falls among a preset number of intervals as a first value, and represent the vector of the intervals that do not fall into as a second value, and obtain the vectors of the preset number of intervals as the vector representation of the cost of each charge item for each patient visit.

[0052] In one embodiment, before step S203, it includes:

[0053] Perform interval division on the costs of all charge items for each patient visit to obtain a preset number of intervals.

[0054] Step S204: Add the representation vectors of the names of the charge items, the representation vectors of the types of the charge items, and the representation vectors of the costs of the charge items for each patient visit to obtain the input representation vector for each patient visit;

[0055] Step S205: Based on the input representation vector for each patient visit, obtain the representation vector of the data for each patient visit.

[0056] In an application, add the representation vectors input into three embedding layers to obtain an input representation vector (input). Input the input representation vector into a deep learning network constructed by a 12-layer machine translation (Transformer) encoder to learn the patient's behavior, and obtain an output representation vector (output). Since the first representation vectors input into the three embedding layers all have special markers, the output representation vector corresponding to the first representation vector can be used as the vector finally representing the patient's single visit (that is, the representation vector of the data for each patient visit (patient representation vector)) and output.

[0057] In one embodiment, step S204 includes the following:

[0058] Input the representation vector of the name of the charge item for each patient visit into a name embedding layer, input the representation vector of the type of the charge item into a type embedding layer, and input the representation vector of the cost of the charge item into a cost embedding layer, so as to add the representation vector of the name of the charge item, the representation vector of the type of the charge item, and the representation vector of the cost of the charge item for each patient visit to obtain the input representation vector for each patient visit. The first representation vectors input into the name embedding layer, the type embedding layer, and the cost embedding layer have special markers;

[0059] Step S205 includes:

[0060] Input the input representation vector of each patient visit into a deep learning network constructed by a 12-layer encoder for patient behavior learning to obtain the output representation vector of each patient visit;

[0061] Use the output representation vector corresponding to the first representation vector in the output representation vector of each patient visit as the representation vector of the data of each patient visit.

[0062] As Figure 3 shown, a schematic structural diagram of a representation learning model for a single patient visit behavior is exemplarily shown; where, N0 to N n respectively represent the name embedding layers of n charging items, T0 to T n respectively represent the type embedding layers of n charging items, C0 to C n respectively represent the cost embedding layers of n charging items, E0 to E n respectively represent the input representation vectors obtained by adding the representation vectors of the names, types, and costs of n charging items after passing through the corresponding embedding layers, Layer1 to Layer12 represent 12-layer machine translation encoders, O0 to O n respectively represent the output representation vectors of each patient visit obtained after the n input representation vectors corresponding to n charging items are respectively learned about patient behavior through 12-layer machine translation encoders, and the representation vector output by O0 is used as the vector finally representing the single patient visit.

[0063] Step S102, based on the representation vector of the data of each patient visit, obtain the representation vector of the data of multiple patient visits of the patient.

[0064] In applications, since the multiple patient visits have time information, the number of inspection and test items for each patient is different and often varies greatly. Therefore, when considering the representation of multiple patient visits, the single vectors cannot be simply concatenated. The representation vectors of multiple patient visits can be fed into a Long Short Term Memory network (Lstm) in chronological order, and after passing through the LSTM, the representation vector of the multiple patient visits of the patient (i.e., the representation vector of the data of multiple patient visits of the patient) is obtained.

[0065] In one embodiment, step S102 includes:

[0066] Input the representation vector of the data of each patient visit into a long short term memory network in chronological order to obtain the representation vector of the data of multiple patient visits of the patient.

[0067] In one embodiment, the expression of the representation vector of the data of multiple patient visits of the patient is:

[0068] E v = LSTM([Embedding TF (v1), Embedding TF (v2), …, Embedding TF (v n )])

[0069] Among them, Embedding TF represents the single-visit behavior learning network, and the single-visit behavior learning network includes the aforementioned three embedding layers and the deep learning network. E v represents the representation vector of the data of the patient's multiple visits. v1, v2, …, v n respectively represent the data of the 1st visit to the nth visit.

[0070] As Figure 4 shown, the structural schematic diagram of the long short-term memory network is exemplarily shown.

[0071] In one embodiment, after step S102, it includes:

[0072] Upload the data of each visit of the patient, the representation vector of the data of each visit of the patient, and the representation vector of the data of the patient's multiple visits to the blockchain.

[0073] In the application, after obtaining any data among the data of each visit of the patient, the representation vector of the data of each visit of the patient, and the representation vector of the data of the patient's multiple visits, the obtained data can be uploaded to the blockchain according to actual needs, which can ensure its security and fairness and transparency to users. The terminal device can download these data from the blockchain to verify whether these data have been tampered with. The blockchain referred to in this example is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm. Blockchain, in essence, is a decentralized database, a string of data blocks generated by using cryptographic methods. Each data block contains information on a batch of network transactions, used to verify the validity of its information (anti-counterfeiting) and generate the next block. The blockchain can include the blockchain underlying platform, the platform product service layer, and the application service layer, etc.

[0074] Step S103: Based on the representation vector of the data of the patient's multiple visits, identify whether the patient is abnormal.

[0075] In the application, after inputting the representation vectors of the data of each visit of the patient into the long short-term memory network for supervised training in chronological order, the representation vector of the data of the patient's multiple visits is obtained, and whether the patient is abnormal can be identified according to the specific value of this representation vector.

[0076] In one embodiment, step S103 includes:

[0077] If the representation vector of the patient's multiple visit data is a third value, determine that the patient is abnormal;

[0078] If the representation vector of the patient's multiple visit data is a fourth value, determine that the patient is normal.

[0079] In application, the third value can specifically be 0, and the fourth value can specifically be 1. That is, a binary vector used as machine language is adopted to represent whether the patient's behavior is abnormal, which is conducive to the terminal device for identification. The user can also set the third value and the fourth value to other values through the man-machine interaction device of the terminal device according to actual needs.

[0080] The abnormal patient identification method provided by this application proposes a unique deep learning network. By fusing multiple embedding layers representing different meanings in the deep learning network structure to represent the patient's visit behavior, and mining the patient's multiple visit behaviors through the deep learning network to identify abnormal patients, it enriches the abnormal patient identification method library.

[0081] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0082] The embodiments of this application also provide an abnormal patient identification device for executing the steps in the above abnormal patient identification method. This device can be a virtual appliance in the terminal device, run by the processor of the terminal device, or the terminal device itself.

[0083] As Figure 5 shown, the abnormal patient identification device 100 provided by the embodiments of this application includes:

[0084] A first vector representation unit 101, configured to perform vector representation on the data of each patient visit to obtain a representation vector of the data of each patient visit;

[0085] A second vector representation unit 102, configured to obtain a representation vector of the data of the patient's multiple visits based on the representation vector of the data of each patient visit;

[0086] An identification unit 103, configured to identify whether the patient is abnormal based on the representation vector of the patient's multiple visit data.

[0087] In one embodiment, the abnormal patient identification device further includes:

[0088] A communication unit, configured to upload data of each patient visit, a representation vector of the data of each patient visit, and a representation vector of the data of multiple patient visits to a blockchain.

[0089] In an application, each unit in the abnormal patient identification device may be a software program unit, or may be implemented by different logic circuits integrated in a processor or independent physical components connected to the processor, or may also be implemented by multiple distributed processors.

[0090] Such as Figure 6 As shown, an embodiment of the present application further provides a terminal device 200, including: at least one processor 201 ( Figure 6 only one processor is shown), a memory 202, and a computer program 203 stored in the memory 202 and executable on at least one processor 201. When the processor 201 executes the computer program 203, the steps in the embodiments of the above various abnormal patient identification methods are implemented.

[0091] In an application, the terminal device may include, but is not limited to, a processor and a memory. Figure 6 This is only an example of the terminal device and does not limit the terminal device. It may include more or fewer components than shown in the figure, or combine some components, or different components. For example, input / output devices, network access devices, etc. It may also include input / output devices, which may include the above-mentioned human-computer interaction devices, communication modules, display screens, etc.

[0092] In an application, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0093] In an application, the memory may be an internal storage unit of a terminal device in some embodiments, such as a hard disk or memory of the terminal device. The memory may also be an external storage device of the terminal device in other embodiments. For example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the terminal device. The memory may also include both the internal storage unit and the external storage device of the terminal device. The memory is used to store an operating system, application programs, a Boot Loader, data, and other programs, such as program codes of computer programs. The memory may also be used to temporarily store data that has been output or will be output.

[0094] In an application, the display screen may be a Thin Film Transistor Liquid Crystal Display (TFT-LCD), a Liquid Crystal Display (LCD), an Organic Electroluminesence Display (OLED), a Quantum Dot Light Emitting Diodes (QLED) display screen, a seven-segment or eight-segment digital tube, etc.

[0095] In an application, the communication module may provide communication solutions for an application on a network device, including Wireless Local Area Networks (WLAN) (such as Wi-Fi networks), Bluetooth, Zigbee, a mobile communication network, a Global Navigation Satellite System (GNSS), Frequency Modulation (FM), Near Field Communication (NFC), Infrared (IR), etc. The communication module may include an antenna. The antenna may have only one element or may be an antenna array including multiple elements. The communication module may receive electromagnetic waves through the antenna, perform frequency modulation and filtering processing on the electromagnetic wave signals, and send the processed signals to the processor. The communication module may also receive a signal to be sent from the processor, perform frequency modulation and amplification on it, and convert it into electromagnetic waves through the antenna and radiate it out.

[0096] It should be noted that for the content such as information interaction and execution process between the above-mentioned devices / units, since it is based on the same concept as the method embodiment of the present application, for its specific functions and the technical effects brought, reference can be specifically made to the method embodiment part, and details will not be elaborated here.

[0097] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the above division of each functional unit is used as an example for illustration. In practical applications, the above functions can be allocated to different functional units according to needs, that is, the internal structure of the device is divided into different functional units to complete all or part of the functions described above. Each functional unit in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working process of the units in the above system can refer to the corresponding process in the foregoing method embodiment, and details will not be elaborated here.

[0098] The embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments for identifying abnormal patients can be implemented.

[0099] The embodiment of the present application provides a computer program product. When the computer program product runs on a terminal device, the terminal device can implement the steps in the above-mentioned method embodiments for identifying abnormal patients.

[0100] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above method embodiments of the present application, a computer program can be used to instruct relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the data acquisition end or the client, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc.

[0101] In the above embodiments, the descriptions of the various embodiments have their own emphases. For parts not detailed or recorded in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0102] Those of ordinary skill in the art will realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0103] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of devices or units can be in electrical, mechanical or other forms.

[0104] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0105] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. And these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application.

Claims

1. A method for identifying abnormal patients, characterized in that, Including: Vectorize the name, type, and cost of each charge item for each patient visit to obtain the representation vectors of the name, type, and cost of each charge item for each patient visit; Input the representation vector of the name of each charge item for each patient visit into the name embedding layer, input the representation vector of the type of the charge item into the type embedding layer, and input the representation vector of the cost of the charge item into the cost embedding layer, so as to add the representation vector of the name of each charge item for each patient visit, the representation vector of the type of the charge item, and the representation vector of the cost of the charge item to obtain the input representation vector for each patient visit. The first representation vectors input into the name embedding layer, the type embedding layer, and the cost embedding layer have special marks; Input the input representation vector for each patient visit into a deep learning network constructed by 12 layers of encoders for patient behavior learning to obtain the output representation vector for each patient visit; Use the output representation vector corresponding to the first representation vector in the output representation vector for each patient visit as the representation vector of the data for each patient visit; Input the representation vectors of the data for each patient visit into a long short-term memory network in chronological order to obtain the representation vectors of the data for multiple patient visits; Based on the representation vectors of the multiple patient visit data, identify whether the patient is abnormal.

2. The abnormal patient identification method according to claim 1, wherein The vectorizing the name, type, and cost of each charge item for each patient visit to obtain the representation vectors of the name, type, and cost of each charge item for each patient visit includes: Vectorize the name of each charge item for each patient visit to obtain the representation vector of the name of each charge item for each patient visit, and the representation vectors of the same name are the same; Vectorize the type of each charge item for each patient visit to obtain the representation vector of the type of each charge item for each patient visit, and the representation vectors of the same type are the same; Vectorize the cost of each charge item for each patient visit to obtain the representation vector of the cost of each charge item for each patient visit, and the representation vectors of the same cost are the same.

3. The abnormal patient identification method according to claim 2, wherein The vectorizing the cost of each charge item for each patient visit to obtain the representation vector of the cost of each charge item for each patient visit includes: Vectorize the intervals in which the cost of each charge item for each patient visit falls into the preset number of intervals as the first value, and vectorize the intervals that do not fall into the preset number of intervals as the second value, to obtain the vectors of the preset number of intervals as the vector representation of the cost of each charge item for each patient visit.

4. The abnormal patient identification method according to claim 3, wherein Before the vectorizing the intervals in which the cost of each charge item for each patient visit falls into the preset number of intervals as the first value, and vectorizing the intervals that do not fall into the preset number of intervals as the second value, includes: Divide the costs of all charge items for each patient visit into intervals to obtain a preset number of intervals.

5. The abnormal patient identification method according to claim 1, wherein, The expression of the representation vector of the data for multiple patient visits is: E v = LSTM([Embedding TF (v1), Embedding TF (v2), …, Embedding TF (v n )]) Among them, Embedding TF represents the single visit behavior learning network, and the single visit behavior learning network includes the name embedding layer, the type embedding layer, the cost embedding layer, and the deep learning network. E v represents the representation vector of the data of the patient's multiple visits, and v1, v2, …, v n respectively represent the data of the first visit to the nth visit.

6. The abnormal patient identification method according to any one of claims 1 to 5, characterized in that, After obtaining the representation vectors of the patient's data for multiple visits based on the representation vectors of the patient's data for each visit, it includes: Upload the data of each patient visit, the representation vectors of the data of each patient visit, and the representation vectors of the data of the patient's multiple visits to the blockchain.

7. An abnormal patient identification device, characterized in that, It includes: A first vector representation unit for respectively performing vector representation on the name, type, and cost of the charged items for each patient visit to obtain the representation vectors of the name, type, and cost of the charged items for each patient visit; input the representation vector of the name of the charged items for each patient visit into the name embedding layer, input the representation vector of the type of the charged items into the type embedding layer, and input the representation vector of the cost of the charged items into the cost embedding layer to add the representation vectors of the name, type, and cost of the charged items for each patient visit to obtain the input representation vector for each patient visit. The first representation vectors input into the name embedding layer, the type embedding layer, and the cost embedding layer have special marks; input the input representation vector for each patient visit into a deep learning network constructed by 12 layers of encoders for patient behavior learning to obtain the output representation vector for each patient visit; use the output representation vector corresponding to the first representation vector in the output representation vector for each patient visit as the representation vector of the data for each patient visit; A second vector representation unit for inputting the representation vectors of the data of each patient visit into a long short-term memory network in chronological order to obtain the representation vectors of the data of the patient's multiple visits; An identification unit for identifying whether the patient is abnormal based on the representation vectors of the patient's multiple visit data.

8. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the abnormal patient identification method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the abnormal patient identification method according to any one of claims 1 to 6.

Images