CN115170335A

Movatterモバイル変換

Info

Publication number: CN115170335A
Application number: CN202210743277.8A
Authority: CN
Inventors: 贺勇; 张顺; 曾震宇
Original assignee: Alibaba Cloud Computing Ltd
Current assignee: Alibaba Cloud Computing Ltd
Priority date: 2022-06-27
Filing date: 2022-06-27
Publication date: 2022-10-11
Anticipated expiration: 2042-06-27
Also published as: CN115170335B

Abstract

The embodiment of the application provides a method and equipment for identifying abnormal treatment records and a storage medium. Taking a patient as an identification unit, and constructing a diagnosis set sequence and a medical item set sequence corresponding to the patient according to the visit time based on a diagnosis set and a medical item set contained in a visit record of the patient within preset time; based on the two sets of sequences, respectively and sequentially calculating the correlation degree of the diagnosis set and the correlation degree of the medical item set between two adjacent visits so as to generate a correlation degree sequence of the diagnosis set and a correlation degree sequence of the medical item set; the two correlation degree sequences can be used as identification bases of abnormal diagnosis records, and if the two correlation degree sequences do not meet preset conditions, the abnormality of the current patient diagnosis record can be determined. In this way, whether the correlation condition among the multiple visits of the patient is abnormal or not can be sensed based on the diagnosis set correlation sequence and the medical item set correlation sequence, and therefore the abnormal visit record can be identified efficiently/accurately.

Description

Translated fromChinese

一种异常就诊记录的识别方法、设备及存储介质A method, device and storage medium for identifying abnormal medical treatment records

技术领域technical field

本申请涉及数据处理技术领域，尤其涉及一种异常就诊记录的识别方法、设备及存储介质。The present application relates to the technical field of data processing, and in particular, to a method, device and storage medium for identifying abnormal medical treatment records.

背景技术Background technique

为了加强医疗保障基金使用监督管理，保障基金安全，促进基金有效使用，维护公民医疗保障合法权益，官方机构制定了《医疗保障基金使用监督管理条例》，且已颁发并施行。医保基金使用主体多、链条长、风险点多、监管难度大，违法行为多发，欺诈骗保问题频发。In order to strengthen the supervision and management of the use of the medical security fund, ensure the safety of the fund, promote the effective use of the fund, and safeguard the legitimate rights and interests of citizens in medical security, the official agency has formulated the "Regulations on the Supervision and Administration of the Use of the Medical Security Fund", which have been issued and implemented. Medical insurance funds have many users, long chains, many risk points, difficult supervision, frequent illegal acts, and frequent fraud and insurance fraud problems.

而在各种欺诈骗保方式中，医保卡借用、伪造病例等行为最为常见，因此需要对该种行为进行重点识别与监管。Among the various fraudulent insurance methods, the borrowing of medical insurance cards and forgery of medical cases are the most common behaviors. Therefore, it is necessary to focus on identification and supervision of such behaviors.

发明内容SUMMARY OF THE INVENTION

本申请的多个方面提供一种异常就诊记录的识别方法、设备及存储介质，用以更加高效/准确地实现异常就诊记录的识别。Various aspects of the present application provide a method, device and storage medium for identifying abnormal medical treatment records, so as to realize the identification of abnormal medical treatment records more efficiently/accurately.

本申请实施例提供一种异常就诊记录的识别方法，包括：The embodiment of the present application provides a method for identifying abnormal medical treatment records, including:

从目标患者在预设时间内的就诊记录中分别抽取诊断集合和医疗项目集合，单次就诊对应一份就诊记录；The diagnosis set and the medical item set are respectively extracted from the medical visit records of the target patient within the preset time, and a single medical visit corresponds to a medical visit record;

按照就诊时间，分别对各次就诊中发生的诊断集合和医疗项目集合进行排序，以获得诊断集合序列和医疗项目集合序列；Sort the diagnosis sets and medical item sets that occurred in each visit according to the visiting time, respectively, to obtain the diagnostic set sequence and the medical item set sequence;

基于所述诊断集合序列，计算任意相邻两次就诊之间的诊断集合相关度，以产生诊断集合相关度序列；Based on the diagnostic set sequence, calculate the diagnostic set correlation between any two adjacent visits to generate a diagnostic set correlation sequence;

基于所述医疗项目集合序列，计算任意相邻两次就诊之间的医疗项目集合相关度；以产生医疗项目集合相关度序列；Based on the medical item set sequence, calculate the medical item set correlation between any two adjacent medical visits; to generate a medical item set correlation sequence;

若所述诊断集合相关度序列和所述医疗项目集合相关度序列不满足预设条件，则确定所述目标患者的就诊记录存在异常。If the correlation degree sequence of the diagnosis set and the correlation degree sequence of the medical item set do not meet a preset condition, it is determined that there is an abnormality in the medical treatment record of the target patient.

本申请实施例还提供一种计算设备，包括存储器和处理器；Embodiments of the present application also provide a computing device, including a memory and a processor;

所述存储器用于存储一条或多条计算机指令；the memory for storing one or more computer instructions;

所述处理器与所述存储器耦合，用于执行所述一条或多条计算机指令，以用于：The processor is coupled to the memory for executing the one or more computer instructions for:

本申请实施例还提供一种存储计算机指令的计算机可读存储介质，当所述计算机指令被一个或多个处理器执行时，致使所述一个或多个处理器执行前述的异常就诊记录的识别方法。Embodiments of the present application further provide a computer-readable storage medium storing computer instructions, which, when executed by one or more processors, cause the one or more processors to perform the aforementioned identification of abnormal medical treatment records method.

在本申请实施例中，以患者作为识别单位，基于患者在预设时间内的就诊记录中包含的诊断集合和医疗项目集合，按照就诊时间构建患者对应的诊断集合序列和医疗项目集合序列；在诊断集合序列和医疗项目集合序列内分别依次计算相邻两次就诊之间的诊断集合相关度和医疗项目集合相关度，以产生诊断集合相关度序列和医疗项目集合相关度序列；这两个相关度序列可作为异常就诊记录的识别依据，若这两个相关度序列不满足预设条件，则可确定当前患者的就诊记录存在异常。这样，可基于前述的诊断集合相关度序列和医疗项目集合相关度序列，来高效/准确地感知患者发生的多次就诊之间是否存在异常的相关性情况，从而高效/准确地实现异常就诊记录的识别。In the embodiment of the present application, the patient is used as the identification unit, and based on the diagnosis set and medical item set included in the patient's medical visit record within the preset time, the diagnosis set sequence and medical item set sequence corresponding to the patient are constructed according to the medical visit time; In the diagnosis set sequence and the medical item set sequence, the diagnostic set correlation degree and the medical item set correlation degree between two consecutive visits are respectively calculated in turn to generate the diagnosis set correlation degree sequence and the medical item set correlation degree sequence; these two correlations The degree sequence can be used as a basis for identifying abnormal medical records. If the two correlation degree sequences do not meet the preset conditions, it can be determined that the current patient's medical record is abnormal. In this way, based on the aforementioned diagnostic set correlation sequence and medical item set correlation sequence, it is possible to efficiently/accurately perceive whether there is an abnormal correlation between multiple visits of a patient, so as to efficiently/accurately realize abnormal medical records. identification.

附图说明Description of drawings

此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：The drawings described herein are used to provide further understanding of the present application and constitute a part of the present application. The schematic embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation of the present application. In the attached image:

图1为本申请一示例性实施例提供的一种异常就诊记录的识别方法的流程示意图；1 is a schematic flowchart of a method for identifying an abnormal medical treatment record provided by an exemplary embodiment of the present application;

图2a为本申请一示例性实施例提供的一种异常就诊记录的识别方法的逻辑示意图；2a is a schematic diagram of a logic diagram of a method for identifying an abnormal medical treatment record provided by an exemplary embodiment of the present application;

图2b为本申请一示例性实施例提供的一种可选实现方式的逻辑示意图；FIG. 2b is a schematic logical diagram of an optional implementation manner provided by an exemplary embodiment of the present application;

图3为本申请一示例性实施例提供的几个患者的相关度序列对应的分析效果示意图；3 is a schematic diagram of the analysis effect corresponding to the correlation sequence of several patients provided by an exemplary embodiment of the present application;

图4为本申请一示例性实施例提供的一种医疗项目集合相关度的计算方案示意图；FIG. 4 is a schematic diagram of a calculation scheme of a medical item set correlation degree provided by an exemplary embodiment of the present application;

图5为本申请一示例性实施例提供的一种计算医疗项目集合之间相关性的实现方式示意图；FIG. 5 is a schematic diagram of an implementation manner of calculating the correlation between medical item sets provided by an exemplary embodiment of the present application;

图6为本申请另一示例性实施例提供的一种计算设备的结构示意图。FIG. 6 is a schematic structural diagram of a computing device according to another exemplary embodiment of the present application.

具体实施方式Detailed ways

为使本申请的目的、技术方案和优点更加清楚，下面将结合本申请具体实施例及相应的附图对本申请技术方案进行清楚、完整地描述。显然，所描述的实施例仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below with reference to the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

目前，通常需要基于大量的、死板的规则来识别异常就诊记录，效率低，准确性与召回率均不足。为此，本申请的一些实施例中：发明人发现，目前的医保卡借用、伪造病例等骗保方式下，为患者购买与消费的医疗项目记录实际上是非真实需求的或者非真实发生的，基于此，提出以患者作为识别单位，基于患者在预设时间内的就诊记录中包含的诊断集合和医疗项目集合，按照就诊时间构建患者对应的诊断集合序列和医疗项目集合序列；在诊断集合序列和医疗项目集合序列内分别依次计算相邻两次就诊之间的诊断集合相关度和医疗项目集合相关度，以产生诊断集合相关度序列和医疗项目集合相关度序列；这两个相关度序列可作为异常就诊记录的识别依据，若这两个相关度序列不满足预设条件，则可确定当前患者的就诊记录存在异常。这样，可基于前述的诊断集合相关度序列和医疗项目集合相关度序列，来高效/准确地感知患者发生的多次就诊之间是否存在异常的相关性情况，从而高效/准确地实现异常就诊记录的识别。At present, it is usually necessary to identify abnormal medical records based on a large number of rigid rules, which has low efficiency and insufficient accuracy and recall rate. For this reason, in some embodiments of this application: the inventor found that under the current insurance fraud methods such as borrowing of medical insurance cards, forging cases, etc., the records of medical items purchased and consumed for patients are actually unreal needs or unreal occurrences, Based on this, it is proposed to use the patient as the identification unit, based on the diagnosis set and medical item set included in the patient's medical visit record within the preset time, to construct the patient's corresponding diagnosis set sequence and medical item set sequence according to the medical visit time; and within the medical item set sequence, respectively calculate the diagnostic set correlation and medical item set correlation between two consecutive visits to generate a diagnostic set correlation sequence and a medical item set correlation sequence; these two correlation sequences can be As a basis for identifying abnormal medical records, if the two correlation sequences do not meet the preset conditions, it can be determined that the current patient's medical records are abnormal. In this way, based on the aforementioned diagnostic set correlation sequence and medical item set correlation sequence, it is possible to efficiently/accurately perceive whether there is an abnormal correlation between multiple visits of a patient, so as to efficiently/accurately realize abnormal medical records. identification.

以下结合附图，详细说明本申请各实施例提供的技术方案。The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

图1为本申请一示例性实施例提供的一种异常就诊记录的识别方法的流程示意图，图2a为本申请一示例性实施例提供的一种异常就诊记录的识别方法的逻辑示意图。该方法可由数据处理装置执行，该数据处理装置可实现为软件和/或硬件的结合，该数据处理装置可集成在计算设备中。参考图1，该方法可包括：FIG. 1 is a schematic flowchart of a method for identifying an abnormal medical treatment record provided by an exemplary embodiment of the present application, and FIG. 2a is a schematic diagram of a logic diagram of a method for identifying an abnormal medical visit record provided by an exemplary embodiment of the present application. The method may be performed by a data processing apparatus, which may be implemented as a combination of software and/or hardware, which may be integrated in a computing device. Referring to Figure 1, the method may include:

步骤100、从目标患者在预设时间内的就诊记录中分别抽取诊断集合和医疗项目集合，单次就诊对应一份就诊记录；Step 100, respectively extracting a diagnosis set and a medical item set from the medical visit records of the target patient within a preset time, and a single medical visit corresponds to a medical visit record;

步骤101、按照就诊时间，分别对各次就诊中发生的诊断集合和医疗项目集合进行排序，以获得诊断集合序列和医疗项目集合序列；Step 101: Sort the diagnosis sets and medical item sets that occurred in each visit to obtain a diagnosis set sequence and a medical item set sequence according to the visiting time;

步骤102、基于诊断集合序列，计算任意相邻两次就诊之间的诊断集合相关度，以产生诊断集合相关度序列；Step 102, based on the diagnostic set sequence, calculate the diagnostic set correlation between any two adjacent medical visits to generate a diagnostic set correlation sequence;

步骤103、基于医疗项目集合序列，计算任意相邻两次就诊之间的医疗项目集合相关度；以产生医疗项目集合相关度序列；Step 103, based on the medical item set sequence, calculate the medical item set correlation degree between any two adjacent medical visits; to generate a medical item set correlation degree sequence;

步骤104、若诊断集合相关度序列和医疗项目集合相关度序列不满足预设条件，则确定目标患者的就诊记录存在异常。Step 104: If the correlation degree sequence of the diagnosis set and the correlation degree sequence of the medical item set do not meet the preset conditions, it is determined that the medical treatment record of the target patient is abnormal.

本实施例提供的异常就诊记录的识别方法可应用于医保基金监督场景、医疗机构行为监管场景等，本实施例对应用场景不做限定。基于本实施例提供的异常就诊记录的识别方法可更加准确、高效地识别出异常就诊记录，从而及时发现医保卡借用、伪造病例等异常行为。本实施例中提及的目标患者可以是发生就诊的任意患者，患者的每次就诊可产生一份就诊记录，就诊记录中通常会按规范记录患者的疾病表现以及当次诊疗情况等信息，也即是，就诊记录中通常存在就诊相关的、丰富的、全面的信息。The method for identifying abnormal medical treatment records provided in this embodiment can be applied to a medical insurance fund supervision scenario, a medical institution behavior supervision scenario, and the like, and this embodiment does not limit the application scenario. Based on the method for identifying abnormal medical treatment records provided in this embodiment, abnormal medical treatment records can be identified more accurately and efficiently, so as to detect abnormal behaviors such as medical insurance card borrowing and forged cases in time. The target patient mentioned in this embodiment can be any patient who has visited a doctor. Each visit of a patient can generate a medical visit record. The medical visit record usually records information such as the patient's disease manifestation and the current diagnosis and treatment situation according to the specifications. That is, there is usually rich and comprehensive information related to the visit in the medical record.

基于此，在步骤100中，可通过各种文本识别技术从目标患者在预设时间内的就诊记录中抽取出其中包含的诊断集合和医疗项目集合。举例来说，可对医保系统中一年内的所有就诊记录按照患者ID进行组织，从而确定每个患者对应的就诊记录并从中抽取出诊断集合和医疗项目集合。其中，诊断集合可用于记录就诊过程中发生的诊断结果；医疗项目集合则可用于记录就诊过程中所消耗的医疗手段，通常可反映在收费明细中。一份就诊记录中可能包含多个诊断集合，例如，主诊断集合、次诊断集合等；还可包含多个医疗项目集合，例如，所使用的药品项目、手术操作项目、护理项目等。Based on this, instep 100, the diagnosis set and the medical item set contained therein can be extracted from the medical visit records of the target patient within the preset time by various text recognition technologies. For example, all medical treatment records within a year in the medical insurance system can be organized according to patient IDs, so as to determine the medical treatment records corresponding to each patient and extract a diagnosis set and a medical item set from it. Among them, the diagnosis set can be used to record the diagnosis results in the process of visiting a doctor; the medical item set can be used to record the medical means consumed in the process of visiting a doctor, which can usually be reflected in the fee details. A medical visit record may contain multiple diagnosis sets, such as primary diagnosis sets, secondary diagnosis sets, etc.; and may also contain multiple medical item sets, such as used drug items, surgical operation items, nursing items, and so on.

参考图1和图2a，在步骤100中，可按照就诊时间，分别对各次就诊中发生的诊断集合和医疗项目集合进行排序，以获得诊断集合序列和医疗项目集合序列。其中，诊断集合序列中的元素为单次就诊中发生的所有诊断集合，医疗项目集合序列中的元素为单次就诊中发生的所有医疗项目集合。Referring to FIG. 1 and FIG. 2a, instep 100, the diagnosis sets and medical item sets occurring in each visit may be sorted according to the time of the visit to obtain the diagnosis set sequence and the medical item set sequence. Wherein, the elements in the sequence of diagnosis collections are all diagnosis collections that occur in a single visit, and the elements in the sequence of medical item collections are the collections of all medical items that occur in a single visit.

本实施例中，在为目标患者构建诊断集合序列和医疗项目集合序列的过程中，可确定各次就诊中分别发生的诊断集合和医疗项目集合；按照就诊时间，对诊断集合进行排序，以获得诊断集合序列；按照就诊时间，对医疗项目集合进行排序，以获得医疗项目集合序列。其中，在诊断集合和医疗项目集合内部，元素(即诊断集合或医疗项目集合)之间可按照发生时间进行排序，相同发生时间的情况下可按照记录的先后顺序进行排序，当然也可以随机排序。另外，考虑到一份就诊记录中包含的诊断集合可能为一个或多个，因此，单次就诊中发生的诊断集合中的元素数量可以是一个或多个，同样，医疗项目集合中的元素数量也可以是一个或多个。In this embodiment, in the process of constructing a diagnosis set sequence and a medical item set sequence for a target patient, the diagnosis sets and medical item sets that occurred in each visit can be determined; the diagnosis sets are sorted according to the visit time to obtain Diagnosis set sequence; sort the medical item set according to the visiting time to obtain the medical item set sequence. Among them, within the diagnosis set and the medical item set, the elements (that is, the diagnosis set or the medical item set) can be sorted according to the occurrence time. If the occurrence time is the same, they can be sorted according to the order of the records, and of course they can also be sorted randomly. . In addition, considering that a diagnosis set contained in a medical visit record may be one or more, the number of elements in the diagnosis set that occurs in a single visit can be one or more. Similarly, the number of elements in the medical item set It can also be one or more.

例如，患者A存在三次就诊，第1份就诊记录中的诊断集合包括：a11，a12，a13，第1份就诊记录中的诊断集合包括：a21，a22，第3份就诊记录中的诊断集合包括：a31，那么患者A的诊断集合序列将为{【a11，a12，a13】，【a21，a22】，【a31】}。其中，【a11，a12，a13】为第1份就诊记录对应的诊断集合。For example, patient A has three visits, the diagnosis set in the first visit record includes: a11, a12, a13, the diagnosis set in the first visit record includes: a21, a22, and the diagnosis set in the third visit record includes: : a31, then the diagnostic set sequence of patient A will be {[a11, a12, a13], [a21, a22], [a31]}. Among them, [a11, a12, a13] is the diagnosis set corresponding to the first medical visit record.

在此基础上，在步骤102和103中，可基于诊断集合序列，计算任意相邻两次就诊之间的诊断集合相关度，以产生诊断集合相关度序列；基于医疗项目集合序列，计算任意相邻两次就诊之间的医疗项目集合相关度；以产生医疗项目集合相关度序列。其中，相邻两次就诊是指时间上相邻的两次就诊。承接上例，患者A的诊断集合序列中包含三次就诊各自对应的诊断集合，其中，【a11，a12，a13】和【a21，a22】相邻，【a21，a22】和【a31】相邻，则可计算【a11，a12，a13】和【a21，a22】之间的相关度以及【a21，a22】和【a31】之间的相关度；这样，针对患者A的诊断集合序列将获得两个诊断集合相关度取值，这两个诊断集合相关度取值可构成患者A对应的诊断集合相关度序列。在诊断集合相关度序列中，元素为相邻就诊之间的诊断集合相关度，另外，元素之间的可按照相关诊断集合在诊断集合序列中的顺序进行排序，当然，这并不是必须的，本实施例对诊断集合相关度序列中的元素顺序不做限定。同理，也可产生患者A的医疗项目集合相关度序列。On this basis, insteps 102 and 103, the diagnostic set correlation between any two adjacent medical visits can be calculated based on the diagnostic set sequence to generate a diagnostic set correlation sequence; Medical item set correlation between two visits; to generate a medical item set correlation sequence. Among them, two consecutive visits refer to two consecutive visits in time. Continuing the above example, the diagnosis set sequence of patient A contains the corresponding diagnosis sets of the three visits. Then the correlation between [a11, a12, a13] and [a21, a22] and between [a21, a22] and [a31] can be calculated; thus, the diagnostic set sequence for patient A will obtain two The diagnosis set correlation degree value, the two diagnosis set correlation degree values may constitute a diagnosis set correlation degree sequence corresponding to patient A. In the diagnostic set correlation sequence, the elements are the diagnostic set correlations between adjacent visits. In addition, the elements can be sorted according to the order of the related diagnostic sets in the diagnostic set sequence. Of course, this is not necessary. This embodiment does not limit the order of elements in the diagnostic set correlation sequence. Similarly, the medical item set correlation degree sequence of patient A can also be generated.

其中，关于诊断集合相关度和医疗项目集合相关度的技术方案将在后文中进行详述。Among them, the technical solutions on the correlation degree of the diagnosis set and the medical item set correlation degree will be described in detail later.

在产生目标患者对应的诊断集合相关度序列和医疗项目集合相关度序列之后，参考图1和图2a，在步骤104中，若诊断集合相关度序列和医疗项目集合相关度序列不满足预设条件，则确定目标患者的就诊记录存在异常。其中，预设条件可以是诊断集合相关度序列和医疗项目集合相关度序列各自对应的相关度代表值不低于预设相关度阈值且各自对应的波动率符合指定波动范围。据此，在步骤104中，可分别计算诊断集合相关度序列和医疗项目集合相关度序列各自对应的相关度代表值，其中，相关度代表值采用相关度均值或中值；还可计算诊断集合相关度序列和医疗项目集合相关度序列各自对应的波动率；若诊断集合相关度序列和医疗项目集合相关度序列各自对应的相关度代表值不低于预设相关度阈值且各自对应的波动率符合指定波动范围，则可确定目标患者未发生异常就诊记录；否则，可确定目标患者的就诊记录存在异常。After generating the diagnosis set correlation degree sequence and medical item set correlation degree sequence corresponding to the target patient, referring to FIG. 1 and FIG. 2a, instep 104, if the diagnosis set correlation degree sequence and the medical item set correlation degree sequence do not meet the preset conditions , it is determined that there is an abnormality in the medical record of the target patient. Wherein, the preset condition may be that the representative value of the correlation degree corresponding to the correlation degree sequence of the diagnosis set and the correlation degree sequence of the medical item set respectively is not lower than the preset correlation degree threshold and the corresponding volatility rates conform to the specified fluctuation range. Accordingly, instep 104, the representative value of the correlation degree corresponding to the correlation degree sequence of the diagnosis set and the correlation degree sequence of the medical item set can be calculated respectively, wherein the representative value of the correlation degree adopts the mean or median value of the correlation degree; the diagnosis set can also be calculated. The respective volatility rates of the correlation sequence and the medical item set correlation sequence; if the respective correlation representative values of the diagnosis set correlation sequence and the medical item set correlation sequence are not lower than the preset correlation threshold and the corresponding volatility If the specified fluctuation range is met, it can be determined that the target patient has no abnormal medical record; otherwise, it can be determined that the target patient's medical record is abnormal.

图3为本申请一示例性实施例提供的几个患者的相关度序列对应的分析效果示意图。参考图3，A图中的序列波动率符合指定波动范围，较为平稳；且序列内的各个元素(相关度)的值都比较高，因此，可确定A图对应的患者未发生异常就诊记录。B图中的序列波动率符合指定波动范围，较为平稳；但是，序列内的各个元素(相关度)的值都比较低，不符合阈值要求，因此，可确定B图对应的患者的就诊记录存在异常。C图中的序列波动率超出了指定波动范围，非常不平稳；且序列内的各个元素(相关度)有高有低，不符合阈值要求，因此，可确定C图对应的患者的就诊记录存在异常。FIG. 3 is a schematic diagram of analysis effects corresponding to correlation degree sequences of several patients provided by an exemplary embodiment of the present application. Referring to Figure 3, the volatility of the sequence in Figure A conforms to the specified fluctuation range and is relatively stable; and the values of each element (correlation) in the sequence are relatively high. Therefore, it can be determined that the patient corresponding to Figure A has no abnormal medical records. The sequence volatility in Figure B conforms to the specified fluctuation range and is relatively stable; however, the values of each element (correlation) in the sequence are relatively low and do not meet the threshold requirements. Therefore, it can be determined that the medical records of the patients corresponding to Figure B exist. abnormal. The volatility of the sequence in Figure C exceeds the specified fluctuation range and is very unstable; and each element (correlation) in the sequence is high or low, which does not meet the threshold requirements. Therefore, it can be determined that the medical records of the patients corresponding to Figure C exist. abnormal.

本实施例中，对异常就诊记录的识别问题进行了技术抽象，对每个患者的就诊记录，按照就诊时间进行排序，分别得到就诊的诊断集合序列、医疗项目集合序列，然后计算相邻的两次就诊之间的诊断集合相关度与医疗项目集合相关度，从而得到相关度序列。而对于一个正常的患者，其相邻两次就诊在时间间隔不长的情况下且非突发性疾病(如外伤)、非流行性疾病(如流感)以及非普遍性疾病(如感冒)的时候，一般是相关的。比如上一次的就诊诊断是高血压，那么本次就诊大概率也是高血压相关的疾病。那么本实施例中认为这种相关性也体现在每次就诊所消耗的诊断集合和医疗项目集合上，也就是两次就诊之间的诊断集合和医疗项目集合也存在高相关性，相关度序列应该是较为平稳的且相关度代表值较高的；当然，正常情况下偶尔出现不相关的情况也是正常的。但是，通常，带有目的性伪造就诊记录的的情况下，伪造的就诊记录中的诊断集合和医疗项目集合变化多端，很难保持伪造的就诊记录与其它正常就诊记录之间的相关性。为此，本实施例中为相关度序列配置了预设条件，通过判断相关度序列是否满足预设条件，即可确定出患者的就诊记录是否存在异常。In this embodiment, the problem of identifying abnormal medical treatment records is technically abstracted, and the medical treatment records of each patient are sorted according to the medical treatment time to obtain the diagnosis collection sequence and medical item collection sequence of the treatment, respectively, and then calculate the adjacent two The correlation degree between the diagnosis set and the medical item set correlation between the second visits is obtained, so as to obtain the correlation degree sequence. For a normal patient, the time interval between two consecutive visits is not long and there are no sudden diseases (such as trauma), non-epidemic diseases (such as influenza) and non-universal diseases (such as colds). time is generally correlated. For example, if the last visit was diagnosed as hypertension, then this visit has a high probability of a hypertension-related disease. Then in this embodiment, it is considered that this correlation is also reflected in the diagnosis set and medical item set consumed by each clinic visit, that is, the diagnosis set and medical item set between two visits also have high correlation, and the correlation sequence It should be relatively stable and the representative value of the correlation is high; of course, it is normal for occasional irrelevant situations under normal circumstances. However, in general, in the case of purposefully forging medical treatment records, the diagnosis set and medical item set in the forged medical treatment records are varied, and it is difficult to maintain the correlation between the forged medical treatment records and other normal medical treatment records. To this end, a preset condition is configured for the correlation sequence in this embodiment, and by judging whether the correlation sequence satisfies the preset condition, it can be determined whether there is an abnormality in the patient's medical treatment record.

综上，本实施例中，以患者作为识别单位，基于患者在预设时间内的就诊记录中包含的诊断集合和医疗项目集合，按照就诊时间构建患者对应的诊断集合序列和医疗项目集合序列；在诊断集合序列和医疗项目集合序列内分别依次计算相邻两次就诊之间的诊断集合相关度和医疗项目集合相关度，以产生诊断集合相关度序列和医疗项目集合相关度序列；这两个相关度序列可作为异常就诊记录的识别依据，若这两个相关度序列不满足预设条件，则可确定当前患者的就诊记录存在异常。这样，可基于前述的诊断集合相关度序列和医疗项目集合相关度序列，来高效/准确地感知患者发生的多次就诊之间是否存在异常的相关性情况，从而高效/准确地实现异常就诊记录的识别。To sum up, in this embodiment, the patient is used as the identification unit, based on the diagnosis set and medical item set included in the patient's medical visit record within the preset time, the diagnosis set sequence and medical item set sequence corresponding to the patient are constructed according to the medical visit time; In the diagnosis set sequence and the medical item set sequence, the diagnostic set correlation degree and the medical item set correlation degree between two consecutive visits are calculated in turn to generate the diagnosis set correlation degree sequence and the medical item set correlation degree sequence; these two The correlation sequence can be used as a basis for identifying abnormal medical records. If the two correlation sequences do not meet the preset conditions, it can be determined that the current patient's medical records are abnormal. In this way, based on the aforementioned diagnostic set correlation sequence and medical item set correlation sequence, it is possible to efficiently/accurately perceive whether there is an abnormal correlation between multiple visits of a patient, so as to efficiently/accurately realize abnormal medical records. identification.

在上述或下述实施例中，可按序遍历医疗项目集合序列中的相邻医疗项目集合；计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度，第一医疗项目集合和第二医疗项目集合分别为目标患者发生的任意相邻两次就诊各自对应的医疗项目集合；继续遍历医疗项目集合序列中其它相邻医疗项目集合，以按序产生医疗项目集合序列中各相邻医疗项目集合之间的相关度而构成医疗项目集合相关度序列。In the above or the following embodiments, adjacent medical item sets in the sequence of medical item sets may be traversed in order; the correlation between the currently traversed first medical item set and the second medical item set is calculated, and the first medical item The set and the second medical item set are respectively the medical item sets corresponding to any two adjacent medical visits of the target patient; continue to traverse other adjacent medical item sets in the medical item set sequence to sequentially generate the medical item sets in the medical item set sequence. The correlation between adjacent medical item sets constitutes a medical item set correlation sequence.

图4为本申请一示例性实施例提供的一种医疗项目集合相关度的计算方案示意图。参考图4，图中的圆形图案代表医疗项目集合序列中的单个医疗项目集合，可依次对目标患者的医疗项目集合序列中相邻的医疗项目集合执行相关度计算操作，从而产生医疗项目集合相关度序列。FIG. 4 is a schematic diagram of a calculation scheme of a medical item set correlation degree provided by an exemplary embodiment of the present application. Referring to FIG. 4, the circular pattern in the figure represents a single medical item set in the medical item set sequence, and the correlation calculation operation can be performed on the adjacent medical item sets in the medical item set sequence of the target patient in turn, thereby generating a medical item set. Correlation sequence.

本实施例中，按序遍历是指按照医疗项目集合序列中的医疗项目集合的时间顺序，依次遍历相邻的医疗项目集合。举例来说，患者B的医疗项目集合序列为{【b11，b12】，【b21，b22，b23】，【b31、b32、b33】，【b41，b42】}，则可依次计算【b11，b12】和【b21，b22，b23】之间的医疗项目集合相关度1，【b21，b22，b23】和【b31、b32、b33】之间的医疗项目集合相关度2，以及【b31、b32、b33】和【b41，b42】之间的医疗项目集合相关度3，这样，可依次获得患者B对应的3个医疗项目集合相关度并构成医疗项目集合相关度序列【医疗项目集合相关度1，医疗项目集合相关度2，医疗项目集合相关度3】。In this embodiment, the in-order traversal refers to sequentially traversing adjacent medical item sets according to the time order of the medical item sets in the medical item set sequence. For example, if the medical item set sequence of patient B is {[b11, b12], [b21, b22, b23], [b31, b32, b33], [b41, b42]}, [b11, b12] can be calculated sequentially ] and [b21, b22, b23] medical item set correlation 1, [b21, b22, b23] and [b31, b32, b33] medical item setcorrelation 2, and [b31, b32, The medical item set correlation degree 3 between b33] and [b41, b42], in this way, the three medical item set correlation degrees corresponding to patient B can be obtained in turn and form the medical item set correlation degree sequence [medical item set correlation degree 1, Medical item setcorrelation degree 2, medical item set correlation degree 3].

本实施例中，可基于医疗项目集合的名称之间的字符串相似度来度量医疗项目集合之间的相关性，但是由于医疗项目集合存在名称多样性的问题，相关的医疗项目集合之间的名称相似度可能并不高，比如“肾衰竭”与“血液透析”两个诊断名称完全不相似，但是实际上这两个诊断是非常相关的。又比如“厄贝沙坦氢氯噻嗪片“与“格列美脲片”这两个治疗高血压的药品，名称字符串是基本不相似的，但是都是治疗高血压的常用药品。因此，根据医疗项目集合之间的名称相似度来度量医疗项目集合之间的相关性可能存在准确性不足的问题。为此，本实施例中提出，可对医疗项目集合进行向量化，从而可从医疗项目集合的实际意义来评估医疗项目集合之间的相关性。In this embodiment, the correlation between the medical item sets can be measured based on the string similarity between the names of the medical item sets. The name similarity may not be high. For example, the two diagnoses "renal failure" and "hemodialysis" are completely dissimilar, but in fact the two diagnoses are very related. Another example is "Irbesartan Hydrochlorothiazide Tablets" and "Glimepiride Tablets", two drugs for the treatment of hypertension, whose name strings are basically different, but they are both commonly used drugs for the treatment of hypertension. Therefore, there may be a problem of insufficient accuracy in measuring the correlation between medical item sets based on the name similarity between medical item sets. Therefore, it is proposed in this embodiment that the medical item set can be vectorized, so that the correlation between the medical item sets can be evaluated from the actual meaning of the medical item set.

图2b为本申请一示例性实施例提供的一种可选实现方式的逻辑视图。参考图2b，本实施例中，可基于全局数据，预先构建各种医疗项目对应的实体向量。FIG. 2b is a logical view of an optional implementation provided by an exemplary embodiment of the present application. Referring to FIG. 2b, in this embodiment, entity vectors corresponding to various medical items may be pre-built based on global data.

在一种示例性方案中，医疗项目的实体向量的构建过程可以是：获取若干患者的就诊记录作为样本，每个患者对应至少一份就诊记录；以就诊为单位，分别构建各份样本中消耗的医疗项目样本序列；将各份样本中消耗的医疗项目样本序列输入向量构建模型，以供向量构建模型捕获医疗项目之间的共存和/或同伴随关系；根据捕获到的共存和/或同伴随关系，将各医疗项目映射至适配的实体向量。In an exemplary solution, the construction process of the entity vector of the medical item may be: obtaining medical records of several patients as samples, each patient corresponding to at least one medical record; taking the medical visit as a unit, constructing the consumption in each sample respectively medical item sample sequence; input the medical item sample sequence consumed in each sample into the vector construction model, so that the vector construction model can capture the coexistence and/or co-accompaniment relationship between medical items; according to the captured coexistence and/or coexistence Along with the relationship, each medical item is mapped to an adapted entity vector.

在该示例性方案中，可将若干患者的就诊记录作为全局数据，例如，可将过去1年时间内的所有就诊记录都作为样本。对于医疗项目集合，采用就诊作为单位，收集每次就诊所消耗的医疗项目而构建出每次就诊对应的医疗项目样本序列，这样，每次就诊可产生一个医疗项目样本序列，每个医疗项目样本序列中的医疗项目可按发生时间进行排序，发生时间相同的情况下可按记录的先后顺序进行排序，或者也随机排序。基于此，可将基于样本而产生的若干医疗项目样本序列输入向量构建模型。本实施例中，向量构建模型可采用word2vec模型、双向编码BERT模型或变换Transformer模型等语言处理模型。由于在该示例性方案中构建了特殊的输入序列，也即是以就诊为单位而构建的医疗项目样本序列，因此，向量构建模型可从这些特殊的输入序列中捕获医疗项目之间的共存和/或同伴随关系。其中，共存关系可以是指医疗项目出现在同一序列中，同伴随关系则可以是指不共现的医疗项目却均与某个/些医疗项目伴随出现。In this exemplary scheme, the visit records of several patients may be used as global data, for example, all visit records within the past 1 year may be used as a sample. For the medical item set, the medical item is used as the unit to collect the medical items consumed by each visit to construct a medical item sample sequence corresponding to each medical visit. In this way, each medical visit can generate a medical item sample sequence, and each medical item sample The medical items in the sequence can be sorted according to the time of occurrence. If the occurrence time is the same, they can be sorted according to the order of the records, or they can also be sorted randomly. Based on this, several medical item sample sequences generated based on the samples can be input into the vector to construct a model. In this embodiment, the vector construction model may adopt a language processing model such as a word2vec model, a bidirectional encoding BERT model, or a transform Transformer model. Since special input sequences are constructed in this exemplary scheme, that is, medical item sample sequences constructed in units of visits, the vector construction model can capture the coexistence and coexistence among medical items from these special input sequences. / or companionship. Wherein, the coexistence relationship may refer to medical items appearing in the same sequence, and the co-accompanying relationship may refer to medical items that do not co-occur but all accompany one/some medical items.

举例来说：for example:

1)X-A-X-B-X出现了很多次(X是可变的0～K个序列元素)，那么医疗项目A与医疗项目B的实体向量相似度会很高，即共现于同一序列。比如“厄贝沙坦氢氯噻嗪片“与“格列美脲片”经常出现在同一个序列中，则“厄贝沙坦氢氯噻嗪片“与“格列美脲片”是非常相关的医疗项目，他们都是治疗高血压的药品。1) X-A-X-B-X appears many times (X is a variable 0-K sequence element), then the entity vector similarity of medical item A and medical item B will be very high, that is, co-occurrence in the same sequence. For example, "Irbesartan Hydrochlorothiazide Tablets" and "Glimepiride Tablets" often appear in the same sequence, then "Irbesartan Hydrochlorothiazide Tablets" and "Glimepiride Tablets" are very related medical items. All medicines are used to treat high blood pressure.

2)X-A-C-X出现了很多次，X-B-C-X也出现了很多次(X是可变的0～K个序列元素)，那么医疗项目A与医疗项目B经常是同一个下文(后面则是下文，前面则是上文)，那么医疗项目A与医疗项目B的实体向量相似度会很高。也即是，医疗项目A和医疗项目B虽然不共现，但两者却均与医疗项目合C伴随出现，因此医疗项目A和医疗项目B之间存在同伴随关系。2) X-A-C-X appears many times, and X-B-C-X also appears many times (X is a variable 0-K sequence element), then medical item A and medical item B are often the same context (the latter is the following, the front is above), then the entity vector similarity between medical item A and medical item B will be very high. That is, although medical item A and medical item B do not co-occur, both of them appear together with medical item C, so there is a co-accompanying relationship between medical item A and medical item B.

这样，实体向量可用于表征医疗项目的实际语义，从而可更加准确、更加合理地表征出医疗项目的实际意义，且实体向量的构建过程中，不仅从单个患者、单次就诊的角度来理解医疗项目的实际意义，而且，还整合了所有患者的全局数据来更加准确地理解医疗项目的真实含义，这可更加准确地表征医疗项目，从而为后续的医疗项目集合相关度计算提供更加准确的依据。In this way, the entity vector can be used to represent the actual semantics of the medical item, so that the actual meaning of the medical item can be more accurately and reasonably represented, and in the construction of the entity vector, the medical treatment is not only understood from the perspective of a single patient and a single visit. The actual meaning of the item, and the global data of all patients is also integrated to more accurately understand the true meaning of the medical item, which can more accurately characterize the medical item, thus providing a more accurate basis for the subsequent calculation of the medical item set correlation .

另外，实体向量之间的相似度(或者表述为距离)与医疗项目之间的共存和/或同伴随关系的强度相关。在该示例方案中，共现和/或同伴随关系的强度更高的医疗项目之间具有更相关的实体向量。In addition, the similarity (or expressed as distance) between entity vectors correlates with the strength of coexistence and/or co-accompanying relationships between medical items. In this example scenario, medical items with higher strengths of co-occurrence and/or concomitant relationships have more correlated entity vectors.

参考图2b，在此基础上，本实施例中，在计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度的过程中，可获取第一医疗项目集合和第二医疗项目集合中各医疗项目各自对应的实体向量；基于医疗项目的实体向量，计算第一医疗项目集合和第二医疗项目集合之间的相关度。这样，在进行异常就医医疗识别的过程中，不再需要实时地计算医疗项目的实体向量，而是可查找预先为各医疗项目映射出的实体向量，以获得第一医疗项目集合和第二医疗项目集合中各医疗项目各自对应的实体向量。Referring to FIG. 2b, on this basis, in this embodiment, in the process of calculating the correlation between the currently traversed first medical item set and the second medical item set, the first medical item set and the second medical item set can be obtained. The entity vector corresponding to each medical item in the item set; the correlation between the first medical item set and the second medical item set is calculated based on the entity vector of the medical item. In this way, in the process of abnormal medical treatment identification, it is no longer necessary to calculate the entity vector of medical items in real time, but to search for the entity vector mapped for each medical item in advance to obtain the first medical item set and the second medical item set. The entity vector corresponding to each medical item in the item set.

本实施例中，可采用多种实现方式来基于医疗项目的实体向量计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度。In this embodiment, various implementations may be adopted to calculate the correlation between the currently traversed first medical item set and the second medical item set based on the entity vector of the medical item.

在一种实现方式中：可分别计算第一医疗项目集合中各医疗项目对应的实体向量和第二医疗项目集合中各医疗项目对应的实体向量之间的相关度；为第一医疗项目集合中的每个医疗项目查找其各自在第二医疗项目集合中所匹配到的最相关医疗项目及对应的最高相关度；为第二医疗项目集合中的每个医疗项目查找其各自在第一医疗项目集合中所匹配到的最相关医疗项目及对应的最高相关度；根据为第一医疗项目集合中各医疗项目查找到的最高相关度，计算第一医疗项目集合对第二医疗项目集合的第一相关度；根据为第二医疗项目集合中各医疗项目查找到的最高相关度，计算第二医疗项目集合对第一医疗项目集合的第二相关度；根据第一相关度和第二相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度。In an implementation manner: the correlation between the entity vector corresponding to each medical item in the first medical item set and the entity vector corresponding to each medical item in the second medical item set can be calculated separately; For each medical item in the second medical item set, find the most relevant medical item matched in the second medical item set and the corresponding highest correlation; for each medical item in the second medical item set The most relevant medical items matched in the set and the corresponding highest correlation; according to the highest correlation found for each medical item in the first medical item set, the first medical item set to the second medical item set is calculated. Correlation; according to the highest correlation found for each medical item in the second medical item set, calculate the second correlation of the second medical item set to the first medical item set; according to the first correlation and the second correlation, A correlation between the first set of medical items and the second set of medical items is determined.

其中，在根据为第一医疗项目集合中各医疗项目查找到的最高相关度，计算第一医疗项目集合对第二医疗项目集合的第一相关度的过程中，可计算为第一医疗项目集合中各医疗项目查找到的最高相关度之间的和值；将和值与第一医疗项目集合的元素个数之间的比值，作为第一相关度；按照同样的方式，可计算出第二相关度。而在根据第一相关度和第二相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度的过程中，则可将第一相关度和第二相关度之间的均值或最大值，作为第一医疗项目集合和第二医疗项目集合之间的相关度。Wherein, in the process of calculating the first correlation between the first medical item set and the second medical item set according to the highest correlation found for each medical item in the first medical item set, the first medical item set can be calculated as The sum value between the highest correlation degrees found by each medical item in relativity. In the process of determining the correlation between the first medical item set and the second medical item set according to the first correlation and the second correlation, the average value between the first correlation and the second correlation may be or the maximum value, as the correlation between the first medical item set and the second medical item set.

举例来说，对于医疗项目集合SetA与医疗项目集合SetB，要评估集合SetB与SetA之间的相关度，那么可基于实体向量，对于SetB中每个元素b从SetA中找到向量最相关的元素a，这样SetB中每个元素从SetA中找到了最相关的元素，然后将这些相关度相加除以SetB的大小便得到了SetB对SetA之间的相关度SimBA，同理对于SetA中每个元素a也从SetB中找到最相关的元素b，这样，SetA中每个元素从SetB中找到了最相关的元素，然后将这些相关度相加处于SetA的大小便得到了SetA对SetB之间的相关度SimAB，最后这两个医疗项目集合的相关度为(SimAB+SimBA)/2或者两者之间的较大值。其中，应当理解的是，若SetB中的元素b1在SetA中的最相关元素为a1，但SetA中的元素a1在SetB中的最相关元素则不一定为b1，因此，在该实现方式中，融合了医疗项目集合之间从不同方向关注到的相关性，可更加准确地度量医疗项目集合之间的相关性。For example, for a medical item set SetA and a medical item set SetB, to evaluate the correlation between the sets SetB and SetA, then based on the entity vector, for each element b in SetB, find the most relevant element a in SetA of the vector , so that each element in SetB finds the most relevant element from SetA, and then adds and divides these correlations by the size of SetB to get the correlation between SetB and SetA SimBA. Similarly, for each element in SetA a also finds the most relevant element b from SetB, so that each element in SetA finds the most relevant element from SetB, and then adds these correlations to the size of SetA to get the correlation between SetA and SetB The degree of SimAB, and finally the correlation between the two medical item sets is (SimAB+SimBA)/2 or the larger value between the two. Among them, it should be understood that if the most relevant element of element b1 in SetB in SetA is a1, but the most relevant element of element a1 in SetA in SetB is not necessarily b1, therefore, in this implementation, The correlation between medical item sets from different directions is integrated, and the correlation between medical item sets can be measured more accurately.

在另一种实现方式中：可以第一医疗项目集合和第二医疗项目集合作为二部图中的两个子集，第一医疗项目集合和第二医疗项目集合中包含的医疗项目作为二部图中的顶点；确定二部图中两个子集之间的顶点连通关系，以获得二部图中包含的多个连通分支；分别在多个连通分支内部计算顶点相关度；根据多个连通分支内部计算出的顶点相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度；其中，若基于实体向量，确定第一医疗项目集合中的第一医疗项目与其在第二医疗项目集合中的最相关第二医疗项目之间的相关度不低于指定阈值，则确定第一医疗项目和第二医疗项目连通。In another implementation manner: the first medical item set and the second medical item set can be used as two subsets in the bipartite graph, and the medical items included in the first medical item set and the second medical item set can be used as the bipartite graph vertices in the bipartite graph; determine the vertex connectivity relationship between the two subsets in the bipartite graph to obtain multiple connected branches contained in the bipartite graph; calculate the vertex correlation within the multiple connected branches respectively; The calculated vertex correlation degree determines the degree of correlation between the first medical item set and the second medical item set; wherein, if based on the entity vector, determine the first medical item in the first medical item set and its in the second medical item set If the correlation between the most relevant second medical items in the set is not lower than the specified threshold, it is determined that the first medical item and the second medical item are connected.

其中，在确定二部图中两个子集之间的顶点连通关系的过程中，可分别计算第一医疗项目集合中各医疗项目对应的实体向量和第二医疗项目集合中各医疗项目对应的实体向量之间的相关度；为第一医疗项目集合中的每个医疗项目查找其各自在第二医疗项目集合中所配到的最相关医疗项目；若对应的最高相关度不低于指定阈值，则可建立相应顶点之间的连通关系；同理，可为第二医疗项目集合中的每个医疗项目查找其各自在第一医疗项目集合中所配到的最相关医疗项目；若对应的最高相关度不低于指定阈值，则可建立相应顶点之间的连通关系。Wherein, in the process of determining the vertex connectivity relationship between the two subsets in the bipartite graph, the entity vector corresponding to each medical item in the first medical item set and the entity corresponding to each medical item in the second medical item set can be calculated respectively. The correlation between the vectors; for each medical item in the first medical item set, find the most relevant medical item assigned in the second medical item set; if the corresponding highest correlation is not lower than the specified threshold, Then the connectivity relationship between the corresponding vertices can be established; in the same way, each medical item in the second medical item set can be searched for the most relevant medical item assigned in the first medical item set; If the correlation is not lower than the specified threshold, the connected relationship between the corresponding vertices can be established.

图5为本申请一示例性实施例提供的一种计算医疗项目集合之间相关性的实现方式示意图。参考图5，在该实现方式中，将医疗项目集合SetA与医疗项目集合SetB作为二部图中的两个子集，对于医疗项目集合SetA与医疗项目集合SetB中的元素(医疗项目集合)则作为二部图中的顶点。在此基础上，可：FIG. 5 is a schematic diagram of an implementation manner of calculating the correlation between medical item sets according to an exemplary embodiment of the present application. Referring to FIG. 5 , in this implementation, the medical item set SetA and the medical item set SetB are used as two subsets in the bipartite graph, and the elements (medical item sets) in the medical item set SetA and the medical item set SetB are used as Vertices in a bipartite graph. On this basis, it is possible to:

1)对于SetA中每一个元素a从SetB中找到最相关元素b，并且相关度不少于阈值sim_threshod，则构造a与b之间的连线；1) For each element a in SetA, find the most relevant element b from SetB, and the correlation is not less than the threshold sim_threshod, then construct the connection between a and b;

2)对于SetB中每一个元素b从SetA中找到最相关元素a，并且相关度不少于阈值sim_threshod，则构造a与b之间的连线；2) For each element b in SetB, find the most relevant element a from SetA, and the correlation is not less than the threshold sim_threshod, then construct the connection between a and b;

3)通过1)与2)得到二部图的所有连通分支，连通分支内的元素之间直接或间接连通。参考图5中，最终得到5个连通分支。3) All connected branches of the bipartite graph are obtained through 1) and 2), and the elements in the connected branches are directly or indirectly connected. Referring to Figure 5, 5 connected branches are finally obtained.

可选地，在多个连通分支内部计算顶点相关度的一种示例性方案可以是：采用连通分支内顶点之间的最大相关度、最小相关度或相关度均值，作为连通分支内部的顶点相关度。而根据多个连通分支内部计算出的顶点相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度的一种示例性方案可以是：计算为多个连通分支对应的顶点相关度之间的和值；将和值与连通分支的数量之间的比值，作为第一医疗项目集合和第二医疗项目集合之间的相关度。Optionally, an exemplary solution for calculating vertex correlations within multiple connected branches may be: using the maximum correlation, minimum correlation or the average correlation between vertices in the connected branch as the vertex correlation within the connected branch. Spend. An exemplary solution for determining the correlation between the first medical item set and the second medical item set according to the vertex correlations calculated inside the multiple connected branches may be: calculating the vertex correlations corresponding to the multiple connected branches The sum value between the degrees; the ratio between the sum value and the number of connected branches is taken as the correlation degree between the first medical item set and the second medical item set.

举例来说，继续参考图5，可在每个连通分支内部计算相关度，多个连通分支的相关度相加除以连通分支的个数得到医疗项目集合SetA与医疗项目集合SetB之间的相关度。其中，连通分支内部计算相关度的计算方法可以使用相关度最大/最小/平均值来表示。假设，连通分支内部采用最大值来表示，图5中共5个连通分支，sim_threshod为0.5。则SetA与SetB的相关度为：(max{0.85}+max{0.9,0.86}+max{0.93,0.95,0.88})/5＝(0.85+0.9+0.95)/5＝0.54。For example, continuing to refer to FIG. 5 , the correlation can be calculated inside each connected branch, and the correlations of multiple connected branches are added and divided by the number of connected branches to obtain the correlation between the medical item set SetA and the medical item set SetB. Spend. Among them, the calculation method of calculating the correlation degree inside the connected branch can be represented by the maximum/minimum/average value of the correlation. Assuming that the connected branch is represented by the maximum value, there are 5 connected branches in Figure 5, and sim_threshod is 0.5. Then the correlation between SetA and SetB is: (max{0.85}+max{0.9,0.86}+max{0.93,0.95,0.88})/5=(0.85+0.9+0.95)/5=0.54.

值得说明的是，上述两种计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度的实现方式进行示例性的，本实施例并不限于此，还可采用其它实现方式来计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度，在此不再穷举。It is worth noting that the above two implementation manners of calculating the correlation between the currently traversed first medical item set and the second medical item set are exemplary, and this embodiment is not limited to this, and other implementation manners may also be used. to calculate the correlation between the currently traversed first medical item set and the second medical item set, which is not exhaustive here.

这样，通过按序遍历目标患者的医疗项目集合序列，可获得能够反映各次就诊之间医疗项目集合相关度情况的医疗项目集合相关度序列。In this way, by traversing the medical item set sequence of the target patient in order, a medical item set correlation degree sequence that can reflect the medical item set correlation degree between each visit can be obtained.

另外，正如前述实施例中提及的，除了可为目标患者构建医疗项目集合序列外，可为目标患者构建诊断集合序列，并基于诊断集合序列产生诊断集合相关度序列。本实施例中，可采用与前述针对医疗项目集合所实施的相同技术方案来产生目标患者对应的诊断集合相关度序列。但是，预先为诊断构建实体向量的实现方案可能与前述的预先为医疗项目构建实体向量的实现方案存在细微差别，主要差别在于，为诊断构建实体向量的过程中，不再以就诊为单位，而是以患者为单位构建样本序列，也即是：可获取若干患者的就诊记录作为样本，每个患者对应至少一份就诊记录；以患者为单位，按照就诊时间，对患者的至少一份就诊记录中出现的诊断集合进行排序，每一份就诊记录中的多个诊断，则按照诊断的医生书写顺序进行排序，或者按照主次排序(主诊断次诊断)，亦或是医生下诊断的时间排序，以产生患者对应的诊断样本序列；将若干患者各自对应的诊断样本序列输入向量构建模型，以供向量构建模型捕获诊断之间的共存和/或同伴随关系；根据捕获到的共存和/或同伴随关系，将各个诊断映射至适配的实体向量。举例来说，患者A三次就诊记录作为样本，第1次就诊包含的诊断集合timeA1：a11，a12，a13，第2次就诊包含的诊断集合timeA2：a21，a22，第3次就诊包含的诊断集合timeA3：a31，那么可为患者A构建诊断样本序列：【a11，a12，a13，a21，a22，a31】。对于患者B两次就诊记录作为样本：第1次就诊包含的诊断集合timeB1：b11，第2次就诊包含的诊断集合timeB2：b21，a22，那么可为患者B构建诊断样本序列：【b11，b21，b22】。这样可构建出每个患者各自对应的诊断样本序列，n个患者就有n个诊断样本序列。基于构建出的这些特殊的序列，可将所有患者各自对应的诊断样本序列一起作为向量构建模型的输入，实体向量可按前文中提及的内部逻辑而输出每个诊断的实体向量。这里，采用患者为单位来为向量构建模型创建样本序列，主要是考虑到单次就诊中发生的诊断数量通常比较少，大部分只有一个诊断，如果还是以就诊作为处理单位，则大部分诊断样本序列将只包含一个元素，将导致实体向量的构建效果不佳。In addition, as mentioned in the foregoing embodiment, in addition to constructing a medical item set sequence for the target patient, a diagnosis set sequence can be constructed for the target patient, and a diagnostic set correlation degree sequence is generated based on the diagnostic set sequence. In this embodiment, the same technical solution as the aforementioned implementation for the medical item set can be used to generate the diagnosis set correlation degree sequence corresponding to the target patient. However, the implementation scheme of constructing entity vectors for diagnosis in advance may be slightly different from the aforementioned implementation scheme of constructing entity vectors for medical items in advance. The sample sequence is constructed in units of patients, that is, the medical records of several patients can be obtained as samples, and each patient corresponds to at least one medical record; the patient is the unit, according to the medical visit time, at least one medical record of the patient The diagnosis sets appearing in the diagnosis set are sorted, and the multiple diagnoses in each medical treatment record are sorted according to the order in which the diagnosis was written by the doctor, or in the order of primary and secondary (primary diagnosis and secondary diagnosis), or the time of the diagnosis made by the doctor. , to generate a sequence of diagnostic samples corresponding to a patient; input the sequence of diagnostic samples corresponding to several patients into a vector construction model, so that the vector construction model captures the coexistence and/or co-accompanying relationship between diagnoses; according to the captured coexistence and/or With the adjoint relationship, each diagnosis is mapped to an adapted entity vector. For example, the records of patient A's three visits are used as samples, the first visit includes the diagnosis set timeA1: a11, a12, a13, the second visit includes the diagnosis set timeA2: a21, a22, the third visit includes the diagnosis set timeA2 timeA3: a31, then a sequence of diagnostic samples can be constructed for patient A: [a11, a12, a13, a21, a22, a31]. For patient B's two visit records as samples: the diagnosis set timeB1: b11 included in the first visit, and the diagnosis set timeB2 included in the second visit: b21, a22, then a diagnostic sample sequence can be constructed for patient B: [b11, b21 , b22]. In this way, a diagnostic sample sequence corresponding to each patient can be constructed, and n patients have n diagnostic sample sequences. Based on these special sequences constructed, the corresponding diagnostic sample sequences of all patients can be used as the input of the vector building model, and the entity vector can output the entity vector of each diagnosis according to the internal logic mentioned above. Here, the unit of patient is used to create a sample sequence for the vector construction model, mainly considering that the number of diagnoses occurring in a single visit is usually relatively small, and most of them have only one diagnosis. The sequence will contain only one element, which will result in poor construction of the entity vector.

为节省篇幅，以下仅简述基于诊断集合序列产生诊断集合相关度序列的过程，相关的技术细节可参考上文中针对医疗项目集合所采用的技术方案的描述，但这不应造成本申请保护范围的损失。In order to save space, the following only briefly describes the process of generating the diagnostic set correlation sequence based on the diagnostic set sequence. For the relevant technical details, please refer to the description of the technical solution adopted for the medical item set above, but this should not cause the protection scope of the present application. Loss.

本实施例中，可按序遍历诊断集合序列中的相邻诊断集合；计算当前遍历的第一诊断集合和第二诊断集合之间的相关度，第一诊断集合和第二诊断集合为目标患者发生的任意相邻两次就诊各自对应的诊断集合；继续遍历诊断集合序列中其它相邻诊断集合，以按序产生诊断集合序列中各相邻诊断集合之间的相关度而构成诊断集合相关度序列。In this embodiment, adjacent diagnosis sets in the sequence of diagnosis collections can be traversed in order; the correlation between the currently traversed first diagnosis collection and the second diagnosis collection is calculated, and the first diagnosis collection and the second diagnosis collection are the target patients Diagnosis sets corresponding to any two adjacent medical visits that occur; continue to traverse other adjacent diagnosis sets in the sequence of diagnosis sets to generate the correlation between adjacent diagnosis sets in the sequence of diagnosis sets in order to form the correlation degree of diagnosis sets sequence.

其中，计算当前遍历的第一诊断集合和第二诊断集合之间的相关度的过程中：可获取第一诊断集合和第二诊断集合中各个诊断各自对应的实体向量，实体向量用于表征诊断的实际语义且共现和/或同伴随关系的强度更高的诊断之间具有更相关的实体向量；基于诊断的实体向量，计算第一诊断集合和第二诊断集合之间的相关度。Wherein, in the process of calculating the correlation between the currently traversed first diagnosis set and the second diagnosis set: the entity vector corresponding to each diagnosis in the first diagnosis set and the second diagnosis set can be obtained, and the entity vector is used to represent the diagnosis Diagnoses with higher actual semantics and co-occurrence and/or higher co-occurrence relationship have more relevant entity vectors; based on the entity vectors of the diagnoses, the correlation between the first diagnosis set and the second diagnosis set is calculated.

在计算第一诊断集合和第二诊断集合之间的相关度的一种实现方式中：可分别计算第一诊断集合中各诊断对应的实体向量和第二诊断集合中各诊断对应的实体向量之间的相关度；为第一诊断集合中的每个诊断查找其各自在第二诊断集合中所匹配到的最相关诊断及对应的最高相关度；为第二诊断集合中的每个诊断查找其各自在第一诊断集合中所匹配到的最相关诊断及对应的最高相关度；根据为第一诊断集合中各诊断查找到的最高相关度，计算第一诊断集合对第二诊断集合的第一相关度；根据为第二诊断集合中各诊断查找到的最高相关度，计算第二诊断集合对第一诊断集合的第二相关度；根据第一相关度和第二相关度，确定第一诊断集合和第二诊断集合之间的相关度。In an implementation manner of calculating the correlation between the first diagnosis set and the second diagnosis set: the difference between the entity vector corresponding to each diagnosis in the first diagnosis set and the entity vector corresponding to each diagnosis in the second diagnosis set may be calculated respectively. The correlation between the two; for each diagnosis in the first diagnosis set, find the most relevant diagnosis and the corresponding highest correlation in the second diagnosis set; for each diagnosis in the second diagnosis set The most relevant diagnosis matched in the first diagnosis set and the corresponding highest correlation; according to the highest correlation found for each diagnosis in the first diagnosis set, the first diagnosis set to the second diagnosis set is calculated. Correlation; according to the highest correlation found for each diagnosis in the second diagnosis set, calculate the second correlation of the second diagnosis set to the first diagnosis set; determine the first diagnosis according to the first correlation and the second correlation The correlation between the set and the second diagnostic set.

其中，根据为第一诊断集合中各诊断查找到的最高相关度，计算第一诊断集合对第二诊断集合的第一相关度的过程中：可计算为第一诊断集合中各诊断查找到的最高相关度之间的和值；将和值与第一诊断集合的元素个数之间的比值，作为第一相关度，同样的方式可计算出第二相关度；根据第一相关度和第二相关度，确定第一诊断集合和第二诊断集合之间的相关度的过程中：可将第一相关度和第二相关度之间的均值或最大值，作为第一诊断集合和第二诊断集合之间的相关度。Wherein, according to the highest correlation found for each diagnosis in the first diagnosis set, in the process of calculating the first correlation degree of the first diagnosis set to the second diagnosis set: it can be calculated for each diagnosis found in the first diagnosis set. The sum value between the highest correlation degrees; the ratio between the sum value and the number of elements in the first diagnostic set is taken as the first correlation degree, and the second correlation degree can be calculated in the same way; Two correlation degrees, in the process of determining the correlation degree between the first diagnosis set and the second diagnosis set: the mean or maximum value between the first correlation degree and the second correlation degree can be used as the first diagnosis set and the second correlation degree. Correlation between diagnostic sets.

在计算第一诊断集合和第二诊断集合之间的相关度的另一种实现方式中：可以第一诊断集合和第二诊断集合作为二部图中的两个子集，第一诊断集合和第二诊断集合中包含的诊断集合作为二部图中的顶点；确定二部图中两个子集之间的顶点连通关系，以获得二部图中包含的多个连通分支；分别在多个连通分支内部计算顶点相关度；根据多个连通分支内部计算出的顶点相关度，确定第一诊断集合和第二诊断集合之间的相关度；其中，若基于实体向量，确定第一诊断集合中的第一诊断与其在第二诊断集合中的最相关第二诊断之间的相关度不低于指定阈值，则确定第一诊断和第二诊断连通。In another implementation manner of calculating the correlation between the first diagnostic set and the second diagnostic set: the first diagnostic set and the second diagnostic set can be used as two subsets in the bipartite graph, the first diagnostic set and the second diagnostic set The diagnostic set included in the bipartite diagnostic set is used as a vertex in the bipartite graph; the vertex connectivity relationship between the two subsets in the bipartite graph is determined to obtain multiple connected branches contained in the bipartite graph; respectively, in the multiple connected branches Calculate the vertex correlation internally; determine the correlation between the first diagnosis set and the second diagnosis set according to the vertex correlation calculated inside the multiple connected branches; wherein, if based on the entity vector, determine the first diagnosis set in the first diagnosis set. The correlation between a diagnosis and its most relevant second diagnosis in the second diagnosis set is not lower than a specified threshold, then it is determined that the first diagnosis and the second diagnosis are connected.

其中，分别在多个连通分支内部计算顶点相关度的过程中：可采用连通分支内顶点之间的最大相关度或相关度均值，作为连通分支内部的顶点相关度；根据多个连通分支内部计算出的顶点相关度，确定第一诊断集合和第二诊断集合之间的相关度的过程中：可计算为多个连通分支对应的顶点相关度之间的和值；将和值与连通分支的数量之间的比值，作为第一诊断集合和第二诊断集合之间的相关度。Among them, in the process of calculating the vertex correlation within the multiple connected branches: the maximum correlation or the average correlation between the vertices in the connected branch can be used as the vertex correlation inside the connected branch; according to the internal calculation of the multiple connected branches In the process of determining the correlation between the first diagnostic set and the second diagnostic set: it can be calculated as the sum value between the vertex correlation degrees corresponding to multiple connected branches; The ratio between the numbers is used as the correlation between the first diagnostic set and the second diagnostic set.

综上，本实施例中，可按需遍历医疗项目集合序列和诊断集合序列，并通过对诊断和医疗项目进行向量化，使用实体向量作为计算相关度的依据，这样，可更加准确地度量各次就诊之间的诊断集合相关度和医疗项目集合相关度，从而获得更加合理、更加准确的诊断集合相关度序列和医疗项目集合相关度序列，进而提供异常就诊记录的识别效率和准确度。To sum up, in this embodiment, the medical item set sequence and the diagnosis set sequence can be traversed as needed, and by vectorizing the diagnosis and medical items, the entity vector can be used as the basis for calculating the correlation, so that each item can be measured more accurately. The diagnostic set correlation and medical item set correlation between two visits can be obtained, so as to obtain a more reasonable and accurate diagnosis set correlation sequence and medical item set correlation sequence, thereby providing the identification efficiency and accuracy of abnormal medical records.

需要说明的是，在上述实施例及附图中的描述的一些流程中，包含了按照特定顺序出现的多个操作，但是应该清楚了解，这些操作可以不按照其在本文中出现的顺序来执行或并行执行，操作的序号如101、102等，仅仅是用于区分开各个不同的操作，序号本身不代表任何的执行顺序。另外，这些流程可以包括更多或更少的操作，并且这些操作可以按顺序执行或并行执行。需要说明的是，本文中的“第一”、“第二”等描述，是用于区分不同的相关度、集合等，不代表先后顺序，也不限定“第一”和“第二”是不同的类型。It should be noted that, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations appearing in a specific order are included, but it should be clearly understood that these operations may not be performed in accordance with the order in which they appear in this document Or in parallel, the sequence numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the sequence numbers themselves do not represent any execution order. Additionally, these flows may include more or fewer operations, and these operations may be performed sequentially or in parallel. It should be noted that the descriptions such as "first" and "second" in this document are used to distinguish different relevancy degrees, sets, etc. different types.

图6为本申请另一示例性实施例提供的一种计算设备的结构示意图。如图6所示，该计算设备包括：存储器60和处理器61。FIG. 6 is a schematic structural diagram of a computing device according to another exemplary embodiment of the present application. As shown in FIG. 6 , the computing device includes: amemory 60 and aprocessor 61 .

处理器61，与存储器60耦合，用于执行存储器60中的计算机程序，以用于：Aprocessor 61, coupled to thememory 60, executes a computer program in thememory 60 for:

基于诊断集合序列，计算任意相邻两次就诊之间的诊断集合相关度，以产生诊断集合相关度序列；Based on the diagnostic set sequence, calculate the diagnostic set correlation between any two adjacent visits to generate a diagnostic set correlation sequence;

基于医疗项目集合序列，计算任意相邻两次就诊之间的医疗项目集合相关度；以产生医疗项目集合相关度序列；Based on the medical item set sequence, calculate the medical item set correlation between any two adjacent visits; to generate the medical item set correlation sequence;

若诊断集合相关度序列和医疗项目集合相关度序列不满足预设条件，则确定目标患者的就诊记录存在异常。If the correlation degree sequence of the diagnosis set and the correlation degree sequence of the medical item set do not meet the preset conditions, it is determined that there is an abnormality in the medical treatment record of the target patient.

在一可选实施例中，处理器61在基于医疗项目集合序列，计算任意相邻两次就诊之间的医疗项目集合相关度；以产生医疗项目集合相关度序列过程中，可用于：In an optional embodiment, theprocessor 61, based on the medical item set sequence, calculates the medical item set correlation between any two adjacent medical visits; in the process of generating the medical item set correlation sequence, it can be used for:

按序遍历医疗项目集合序列中的相邻医疗项目集合；Traverse the adjacent medical item sets in the sequence of medical item sets in order;

计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度，第一医疗项目集合和第二医疗项目集合分别为目标患者发生的任意相邻两次就诊各自对应的医疗项目集合；Calculate the correlation between the currently traversed first medical item set and the second medical item set, where the first medical item set and the second medical item set are the respective medical item sets corresponding to any two adjacent visits to the target patient. ;

继续遍历医疗项目集合序列中其它相邻医疗项目集合，以按序产生医疗项目集合序列中各相邻医疗项目集合之间的相关度而构成医疗项目集合相关度序列。Continue to traverse other adjacent medical item sets in the medical item set sequence to sequentially generate the correlations between the adjacent medical item sets in the medical item set sequence to form a medical item set correlation degree sequence.

在一可选实施例中，处理器61在计算当前遍历的第一医疗项目集合和第二医疗项目集合之间的相关度过程中，可用于：In an optional embodiment, in the process of calculating the correlation between the currently traversed first medical item set and the second medical item set, theprocessor 61 may be used for:

获取第一医疗项目集合和第二医疗项目集合中各个医疗项目各自对应的实体向量，医疗项目对应的实体向量用于表征医疗项目的实际语义且共现和/或同伴随关系的强度更高的医疗项目之间具有更相关的实体向量；Obtain the entity vector corresponding to each medical item in the first medical item set and the second medical item set. The entity vector corresponding to the medical item is used to represent the actual semantics of the medical item and the co-occurrence and/or the strength of the co-accompanying relationship is higher. There are more related entity vectors between medical items;

基于医疗项目的实体向量，计算第一医疗项目集合和第二医疗项目集合之间的相关度。Based on the entity vector of medical items, the degree of correlation between the first medical item set and the second medical item set is calculated.

在一可选实施例中，处理器61在基于医疗项目的实体向量，计算第一医疗项目集合和第二医疗项目集合之间的相关度过程中，可用于：In an optional embodiment, in the process of calculating the correlation between the first medical item set and the second medical item set based on the entity vector of the medical item, theprocessor 61 may be used for:

分别计算第一医疗项目集合中各医疗项目对应的实体向量和第二医疗项目集合中各医疗项目对应的实体向量之间的相关度；respectively calculating the correlation between the entity vector corresponding to each medical item in the first medical item set and the entity vector corresponding to each medical item in the second medical item set;

为第一医疗项目集合中的每个医疗项目查找其各自在第二医疗项目集合中所匹配到的最相关医疗项目及对应的最高相关度；Searching for each medical item in the first medical item set the most relevant medical item matched in the second medical item set and the corresponding highest correlation;

为第二医疗项目集合中的每个医疗项目查找其各自在第一医疗项目集合中所匹配到的最相关医疗项目及对应的最高相关度；Searching for each medical item in the second medical item set the most relevant medical item matched in the first medical item set and the corresponding highest correlation;

根据为第一医疗项目集合中各医疗项目查找到的最高相关度，计算第一医疗项目集合对第二医疗项目集合的第一相关度；Calculate the first correlation of the first medical item set to the second medical item set according to the highest correlation found for each medical item in the first medical item set;

根据为第二医疗项目集合中各医疗项目查找到的最高相关度，计算第二医疗项目集合对第一医疗项目集合的第二相关度；According to the highest correlation found for each medical item in the second medical item set, calculate the second correlation of the second medical item set to the first medical item set;

根据第一相关度和第二相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度。According to the first correlation degree and the second correlation degree, the correlation degree between the first medical item set and the second medical item set is determined.

在一可选实施例中，处理器61在根据为第一医疗项目集合中各医疗项目查找到的最高相关度，计算第一医疗项目集合对第二医疗项目集合的第一相关度过程中，可用于：In an optional embodiment, theprocessor 61 calculates the first correlation degree of the first medical item set to the second medical item set according to the highest correlation found for each medical item in the first medical item set, can be use on:

计算为第一医疗项目集合中各医疗项目查找到的最高相关度之间的和值；Calculate the sum of the highest correlations found for each medical item in the first medical item set;

将和值与第一医疗项目集合的规格之间的比值，作为第一相关度；Taking the ratio between the sum value and the specification of the first medical item set as the first correlation;

根据第一相关度和第二相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度过程中，可用于：According to the first correlation degree and the second correlation degree, in the process of determining the correlation degree between the first medical item set and the second medical item set, it can be used for:

将第一相关度和第二相关度之间的均值或最大值，作为第一医疗项目集合和第二医疗项目集合之间的相关度。The mean or maximum value between the first correlation degree and the second correlation degree is taken as the correlation degree between the first medical item set and the second medical item set.

以第一医疗项目集合和第二医疗项目集合作为二部图中的两个子集，第一医疗项目集合和第二医疗项目集合中包含的医疗项目作为二部图中的顶点；Taking the first medical item set and the second medical item set as two subsets in the bipartite graph, and the medical items included in the first medical item set and the second medical item set as vertices in the bipartite graph;

确定二部图中两个子集之间的顶点连通关系，以获得二部图中包含的多个连通分支；Determine the vertex connectivity relationship between two subsets in the bipartite graph to obtain multiple connected branches contained in the bipartite graph;

分别在多个连通分支内部计算顶点相关度；Calculate vertex correlations within multiple connected branches respectively;

根据多个连通分支内部计算出的顶点相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度；Determine the correlation between the first medical item set and the second medical item set according to the vertex correlations calculated inside the plurality of connected branches;

其中，若基于实体向量，确定第一医疗项目集合中的第一医疗项目与其在第二医疗项目集合中的最相关第二医疗项目之间的相关度不低于指定阈值，则确定第一医疗项目和第二医疗项目连通。Wherein, if it is determined based on the entity vector that the correlation between the first medical item in the first medical item set and the most relevant second medical item in the second medical item set is not lower than the specified threshold, then the first medical item is determined The project is connected with the second medical project.

在一可选实施例中，处理器61在分别在多个连通分支内部计算顶点相关度过程中，可用于：采用连通分支内顶点之间的最大相关度或相关度均值，作为连通分支内部的顶点相关度；In an optional embodiment, theprocessor 61 may be configured to: use the maximum correlation degree or the average correlation degree between vertices in the connected branches, as the correlation degree inside the connected branch, in the process of calculating the vertex correlation degree in the plurality of connected branches respectively. vertex correlation;

根据多个连通分支内部计算出的顶点相关度，确定第一医疗项目集合和第二医疗项目集合之间的相关度过程中，可用于：In the process of determining the correlation between the first medical item set and the second medical item set according to the vertex correlations calculated inside the multiple connected branches, it can be used for:

计算为多个连通分支对应的顶点相关度之间的和值；Calculated as the sum of the vertex correlations corresponding to multiple connected branches;

将和值与连通分支的数量之间的比值，作为第一医疗项目集合和第二医疗项目集合之间的相关度。The ratio between the sum value and the number of connected branches is taken as the correlation between the first medical item set and the second medical item set.

在一可选实施例中，处理器61在医疗项目的实体向量的构建过程，可用于：In an optional embodiment, theprocessor 61 can be used to:

获取若干患者的就诊记录作为样本，每个患者对应至少一份就诊记录；Obtain the medical records of several patients as samples, and each patient corresponds to at least one medical record;

以就诊为单位，分别构建各份样本中消耗的医疗项目样本序列；Construct the sample sequence of medical items consumed in each sample by taking the visit as a unit;

将各份样本中消耗的医疗项目样本序列输入向量构建模型，以供向量构建模型捕获医疗项目之间的共存和/或同伴随关系；Input the medical item sample sequence consumed in each sample into the vector construction model, so that the vector construction model captures the coexistence and/or co-accompaniment relationship between medical items;

根据捕获到的共存和/或同伴随关系，将各医疗项目映射至适配的实体向量。Each medical item is mapped to an adapted entity vector according to the captured coexistence and/or co-occurrence relationship.

在一可选实施例中，处理器61在获取第一医疗项目集合和第二医疗项目集合中各医疗项目各自对应的实体向量过程中，可用于：In an optional embodiment, in the process of acquiring the entity vectors corresponding to each medical item in the first medical item set and the second medical item set, theprocessor 61 may be used for:

查找预先为各医疗项目映射出的实体向量，以获得第一医疗项目集合和第二医疗项目集合中各医疗项目各自对应的实体向量。The entity vector mapped for each medical item in advance is searched to obtain the entity vector corresponding to each medical item in the first medical item set and the second medical item set.

在一可选实施例中，向量构建模型采用word2vec模型、双向编码BERT模型或变换Transformer模型。In an optional embodiment, the vector construction model adopts the word2vec model, the bidirectional encoding BERT model or the transform Transformer model.

在一可选实施例中，处理器61在基于诊断集合序列，计算任意相邻两次就诊之间的诊断集合相关度；以产生诊断集合相关度序列过程中，可用于：In an optional embodiment, theprocessor 61, in the process of calculating the diagnostic set correlation between any two adjacent medical visits based on the diagnostic set sequence; in the process of generating the diagnostic set correlation sequence, can be used for:

按序遍历诊断集合序列中的相邻诊断集合；Traverse adjacent diagnostic sets in the sequence of diagnostic sets in order;

基于当前遍历的第一诊断集合和第二诊断集合中各个诊断各自对应的实体向量，计算第一诊断集合和第二诊断集合之间的相关度，第一诊断集合和第二诊断集合为目标患者发生的任意相邻两次就诊各自对应的诊断集合，诊断对应的实体向量用于表征诊断的实际语义且共现和/或同伴随关系的强度更高的诊断之间具有更相关的实体向量；Based on the entity vector corresponding to each diagnosis in the currently traversed first diagnosis set and the second diagnosis set, the correlation between the first diagnosis set and the second diagnosis set is calculated, and the first diagnosis set and the second diagnosis set are the target patients The corresponding diagnosis sets of any two adjacent medical visits that occur, the entity vector corresponding to the diagnosis is used to represent the actual semantics of the diagnosis, and the co-occurrence and/or the diagnosis with the higher strength of the concomitant relationship has a more relevant entity vector;

继续遍历诊断集合序列中其它相邻诊断集合，以按序产生诊断集合序列中各相邻诊断集合之间的相关度而构成诊断集合相关度序列。Continue to traverse other adjacent diagnosis sets in the sequence of diagnosis sets to sequentially generate the correlations between the adjacent diagnosis sets in the sequence of diagnosis sets to form a sequence of correlation degrees of diagnosis sets.

在一可选实施例中，处理器61在诊断的实体向量的构建过程，可用于：In an optional embodiment, in the process of constructing the entity vector of diagnosis, theprocessor 61 can be used for:

以患者为单位，按照就诊时间，对患者的至少一份就诊记录中出现的诊断集合进行排序，以产生患者对应的诊断样本序列；Taking the patient as a unit and according to the visit time, sort the diagnosis sets appearing in at least one medical visit record of the patient to generate a sequence of diagnosis samples corresponding to the patient;

将若干患者各自对应的诊断样本序列输入向量构建模型，以供向量构建模型捕获诊断之间的共存和/或同伴随关系；Inputting the respective corresponding diagnostic sample sequences of several patients into the vector construction model, so that the vector construction model captures the coexistence and/or co-accompaniment relationship between diagnoses;

根据捕获到的共存和/或同伴随关系，将各个诊断映射至适配的实体向量。Based on the captured coexistence and/or co-occurrence relationships, each diagnosis is mapped to an adapted entity vector.

在一可选实施例中，预设条件包括诊断集合相关度序列和医疗项目集合相关度序列各自对应的相关度代表值不低于预设相关度阈值且各自对应的波动率符合指定波动范围；In an optional embodiment, the preset conditions include that the respective representative values of the correlation degree corresponding to the correlation degree sequence of the diagnosis set and the medical item set correlation degree sequence are not lower than the preset correlation degree threshold and the corresponding volatility rates are within the specified fluctuation range;

相关度代表值采用相关度均值或中值。The representative value of the correlation degree is the mean or median value of the correlation degree.

进一步，如图6所示，该计算设备还包括：通信组件62、电源组件63等其它组件。图6中仅示意性给出部分组件，并不意味着计算设备只包括图6所示组件。Further, as shown in FIG. 6 , the computing device further includes: acommunication component 62 , apower supply component 63 and other components. Only some components are schematically shown in FIG. 6 , which does not mean that the computing device only includes the components shown in FIG. 6 .

值得说明的是，上述关于计算设备各实施例中的技术细节，可参考前述的方法实施例中的相关描述，为节省篇幅，在此不再赘述，但这不应造成本申请保护范围的损失。It is worth noting that, for the technical details in the above-mentioned embodiments of the computing device, reference may be made to the relevant descriptions in the foregoing method embodiments. To save space, details are not repeated here, but this should not cause any loss of the protection scope of the present application. .

相应地，本申请实施例还提供一种存储有计算机程序的计算机可读存储介质，计算机程序被执行时能够实现上述方法实施例中可由计算设备执行的各步骤。Correspondingly, the embodiments of the present application further provide a computer-readable storage medium storing a computer program, and when the computer program is executed, each step that can be executed by a computing device in the foregoing method embodiments can be implemented.

上述图6中的存储器，用于存储计算机程序，并可被配置为存储其它各种数据以支持在计算平台上的操作。这些数据的示例包括用于在计算平台上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。The memory in FIG. 6 described above is used to store computer programs and may be configured to store various other data to support operations on the computing platform. Examples of such data include instructions for any application or method operating on the computing platform, contact data, phonebook data, messages, pictures, videos, etc. Memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

上述图6中的通信组件，被配置为便于通信组件所在设备和其他设备之间有线或无线方式的通信。通信组件所在设备可以接入基于通信标准的无线网络，如WiFi，2G、3G、4G/LTE、6G等移动通信网络，或它们的组合。在一个示例性实施例中，通信组件经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。The above-mentioned communication component in FIG. 6 is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi, a mobile communication network such as 2G, 3G, 4G/LTE, 6G, or a combination thereof. In one exemplary embodiment, the communication component receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication assembly further includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

上述图6中的电源组件，为电源组件所在设备的各种组件提供电力。电源组件可以包括电源管理系统，一个或多个电源，及其他与为电源组件所在设备生成、管理和分配电力相关联的组件。The power supply assembly in FIG. 6 above provides power for various components of the equipment where the power supply assembly is located. A power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the equipment in which the power supply assembly is located.

本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本申请可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。As will be appreciated by those skilled in the art, the embodiments of the present application may be provided as a method, a system, or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

本申请是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flows of the flowcharts and/or the block or blocks of the block diagrams.

这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。Memory may include non-persistent memory in computer readable media, random access memory (RAM) and/or non-volatile memory in the form of, for example, read only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带式磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。Computer-readable media includes both persistent and non-permanent, removable and non-removable media, and storage of information may be implemented by any method or technology. Information may be computer readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), Flash Memory or other memory technology, Compact Disc Read Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, Magnetic tape cartridges, tape-based disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

还需要说明的是，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、商品或者设备不仅包括那些要素，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、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。It should also be noted that the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device comprising a series of elements includes not only those elements, but also Other elements not expressly listed, or which are inherent to such a process, method, article of manufacture, or apparatus are also included. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, method, article of manufacture, or device that includes the element.

以上所述仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的保护范围之内。The above descriptions are merely examples of the present application, and are not intended to limit the present application. Various modifications and variations of this application are possible for those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included within the protection scope of this application.