Non-professional service personnel-oriented reproductive diagnosis and treatment patient portrait generation method and device

Abstract

The invention discloses a non-professional service personnel-oriented reproductive diagnosis and treatment patient portrait generation method and device, and a server side method comprises the steps: regularly obtaining and preprocessing original medical data of each reproductive diagnosis and treatment patient through a data interface provided by an HIS system / LIS system, and obtaining a desensitization structured data set of each reproductive diagnosis and treatment patient; according to the desensitization structured data set, constructing treatment stage parameters and emotion quantization parameters of each reproductive diagnosis and treatment patient; matching a plurality of target portrait tags belonging to the treatment stage parameters and the emotion quantization parameters from a preset portrait tag library; and filling the multiple target portrait labels into a preset natural language template to obtain a portrait description text of each reproductive diagnosis and treatment patient, and sending the portrait description text to a client of non-professional service personnel for display. Therefore, according to the embodiment of the invention, non-professional service personnel can quickly obtain the key information, the service efficiency is improved, the communication quality with the patient is improved, and the satisfaction degree of the patient is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart medical treatment, and in particular to a patient portrait generation method and device for non-professional service personnel in reproductive diagnosis and treatment. BACKGROUND

[0002] In the field of modern reproductive diagnosis and treatment, patients often need to go through multiple complex treatment stages, such as ovulation induction, egg retrieval, embryo culture and transplantation, etc. These treatment processes involve a large amount of medical data and professional terms. When communicating with patients, medical personnel need to rely on the portrait system to quickly understand the core information of the patients in order to provide effective services and support.

[0003] At present, the existing patient portrait system mainly relies on data in electronic medical records (EMR), laboratory information systems (LIS) and hospital information systems (HIS). These systems generate patient portraits by integrating patient diagnosis and treatment records, test reports and appointment history, etc.

[0004] However, the existing patient portrait system mainly serves the clinical decision of expert doctors and relies on a large number of professional terms. These terms are difficult for non-professional service personnel (such as hospital customer service and nurses) to understand, and it is difficult for them to understand at a glance. Before communication, they must check the information or repeatedly ask the doctor, which affects the service efficiency when they communicate with the patients. SUMMARY

[0005] The embodiments of the present application provide a patient portrait generation method and device for non-professional service personnel in reproductive diagnosis and treatment. In order to have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This part is not a general review, nor does it determine the key / important elements or describe the protection scope of these embodiments. The only purpose is to present some concepts in a simple form as a preface to the detailed description below.

[0006] In a first aspect, the embodiments of the present application provide a patient portrait generation method for non-professional service personnel in reproductive diagnosis and treatment, applied to a service end, the method comprising: obtaining and preprocessing the original medical data of each reproductive diagnosis and treatment patient through the data interface provided by the HIS system / LIS system to obtain the desensitization structured data set of each reproductive diagnosis and treatment patient; constructing the treatment stage parameters and emotional quantization parameters of each reproductive diagnosis and treatment patient according to the desensitization structured data set; matching a plurality of target portrait labels belonging to the treatment stage parameters and emotional quantization parameters from a preset portrait label library; Fill the multiple target image labels into the preset natural language template to obtain the image description text of each reproductive diagnosis and treatment patient, and send the image description text to the client of the non-professional service personnel for display.

[0007] Optionally, according to the desensitization structured data set, a treatment stage parameter and an emotion quantization parameter of each reproductive diagnosis and treatment patient are constructed, including: According to the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph and the desensitization structured data set, the current treatment stage of each reproductive diagnosis and treatment patient is determined. According to the desensitization structured data set, the remaining treatment days of each reproductive diagnosis and treatment patient are predicted. The current treatment stage and the remaining treatment days are taken as the treatment stage parameter of each reproductive diagnosis and treatment patient. According to the desensitization structured data set, the communication preference and the psychological state of each reproductive diagnosis and treatment patient are analyzed as the emotion quantization parameter of each reproductive diagnosis and treatment patient.

[0008] Optionally, the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph carries multiple clinical attributes, and each treatment path node is sequentially connected according to the order of the ovulation induction stage, the oocyte retrieval stage, the embryo culture stage, the embryo transfer stage and the follow-up stage. According to the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph and the desensitization structured data set, the current treatment stage of each reproductive diagnosis and treatment patient is determined, including: From the preset reproductive diagnosis and treatment knowledge graph, the clinical feature attributes and conditions carried by each defined treatment path node are obtained. The first structured data between each treatment path node and the clinical feature attributes and conditions carried by each treatment path node is established. From the desensitization structured data set, multiple field attributes related to each treatment path node are extracted as multiple actual field attributes of each treatment path node. The second structured data between each treatment path node and the multiple actual field attributes of each treatment path node is established. According to the order of the preset reproductive diagnosis and treatment knowledge graph, each target node is traversed to be combined into a target node sequence. According to the target node sequence, the first structured data and the second structured data, the current treatment stage of each reproductive diagnosis and treatment patient is determined.

[0009] Optionally, the position index of each target node in the target node sequence is unique. According to the target node sequence, the first structured data and the second structured data, the current treatment stage of each reproductive diagnosis and treatment patient is determined, including: For each target node in the target node sequence, obtain the clinical feature attribute and condition from the first structured data; For each target node in the target node sequence, obtain a plurality of actual field attributes from the second structured data; Match the clinical feature attribute and condition of each target node with the plurality of actual field attributes to determine whether the clinical feature attribute and condition exist in the plurality of actual field attributes, and obtain the determination result of each target node in the target node sequence; Determine the target node indicated by the last determination result as the current treatment stage of each reproductive diagnosis and treatment patient.

[0010] Optionally, the preset reproductive diagnosis and treatment knowledge graph is generated according to the following steps, comprising: Read the reproductive specialist medical record template and clinical guidelines; Define the node path of the ovulation induction stage, oocyte retrieval stage, embryo culture stage, embryo transfer stage, and follow-up stage as the main node path of the graph through the reproductive specialist medical record template and clinical guidelines; Traverse the historical anonymous cases to count the clinical feature attributes and conditions related to each main node path of the graph; Label the clinical feature attributes and conditions corresponding to each main node path of the graph on each main node path of the graph to obtain the preset reproductive diagnosis and treatment knowledge graph.

[0011] Optionally, according to the desensitized structured data set, the remaining treatment days of each reproductive diagnosis and treatment patient are predicted, comprising: Determine a plurality of historical completed treatment stages before the current treatment stage; Obtain the start date and end date of each historical completed treatment stage from the desensitized structured data set; Calculate the difference between the start date and the end date of each historical completed treatment stage to obtain the historical treatment period of each historical completed treatment stage; Calculate the mean and standard deviation of the historical completed treatment stage according to the historical treatment period of each historical completed treatment stage; Calculate the remaining treatment days of each reproductive diagnosis and treatment patient according to the average time consumption and the standard deviation; The calculation formula of the remaining treatment days of each reproductive diagnosis and treatment patient is:

[0012] wherein, is an adjustment coefficient, which is adjusted according to clinical experience and data distribution, is the average time consumption, is the standard deviation; Average time The calculation formula is:

[0013] in, is the number of historically completed treatment phases, It is historical treatment cycles with completed treatment phases; The formula for calculating the standard deviation is:

[0014] in, is the standard deviation.

[0015] Optionally, analyze the communication preferences and psychological state of each reproductive diagnosis and treatment patient based on the desensitized structured data set, including: Extract the communication records and psychological status-related field attributes of each reproductive diagnosis and treatment patient from the desensitized structured data set; Based on the communication records, the frequency of each preset communication method for each reproductive diagnosis and treatment patient was counted; The preset communication method that appears most frequently is used as the communication preference of each reproductive diagnosis and treatment patient; Convert the psychological state related field attributes into emotion description text to obtain the emotion description text of each reproductive diagnosis and treatment patient; Input the emotional description text of each reproductive diagnosis and treatment patient into the pre-trained psychological state recognition model, and output the psychological state of each reproductive diagnosis and treatment patient; The following steps are used to generate a pre-trained mental state recognition model, including: Collect multiple historical emotional description texts for each historical reproductive diagnosis and treatment patient; Label each historical emotion description text with a psychological state label to obtain model training samples; A neural network is used to construct a mental state recognition model, and model training samples are used to perform machine learning on the mental state recognition model to obtain a pre-trained mental state recognition model.

[0016] Optionally, the treatment stage parameters include the current treatment stage and the number of remaining treatment days; the emotional quantification parameters include communication preferences and psychological state; the preset portrait tag library includes each portrait tag and the portrait parameter conditions corresponding to each portrait tag, and the portrait parameter conditions include treatment stage conditions and emotional quantification conditions; From the preset portrait tag library, match multiple target portrait tags belonging to treatment stage parameters and emotional quantitative parameters, including: Traverse each target portrait tag in the preset portrait tag library; analyze whether the current treatment stage and the remaining treatment days meet the treatment stage condition corresponding to each target portrait label in the traversal; analyze whether the communication preference and the psychological state meet the emotional quantification condition corresponding to each target portrait label in the traversal; cache the target portrait labels that meet the treatment stage condition and the emotional quantification condition as the target portrait labels belonging to the treatment stage parameter and the emotional quantification parameter.

[0017] Optionally, the original medical data of each reproductive diagnosis and treatment patient is obtained and preprocessed to obtain a desensitized structured data set of each reproductive diagnosis and treatment patient, including: obtaining the original medical data of each reproductive diagnosis and treatment patient; extracting a plurality of key field information related to reproductive diagnosis and treatment from the original medical data; calling a TF-IDF sensitive word recognition model and loading a reproductive medical exclusive dictionary; desensitizing the plurality of key field information according to the TF-IDF sensitive word recognition model and the reproductive medical exclusive dictionary to obtain the desensitized structured data set of each reproductive diagnosis and treatment patient.

[0018] In a second aspect, the embodiments of the present application provide a reproductive diagnosis and treatment patient portrait generation device for non-professional service personnel, the device comprising: a data preprocessing module configured to periodically obtain and preprocess the original medical data of each reproductive diagnosis and treatment patient through a data interface provided by a HIS system / LIS system to obtain a desensitized structured data set of each reproductive diagnosis and treatment patient; a parameter construction module configured to construct a treatment stage parameter and an emotional quantification parameter of each reproductive diagnosis and treatment patient according to the desensitized structured data set; a portrait label matching module configured to match a plurality of target portrait labels belonging to the treatment stage parameter and the emotional quantification parameter from a preset portrait label library; a portrait description text display module configured to fill the plurality of target portrait labels into a preset natural language template to obtain a portrait description text of each reproductive diagnosis and treatment patient, and send the portrait description text to a client of the non-professional service personnel for display.

[0019] The technical scheme provided by the embodiments of the present application can include the following beneficial effects: In the embodiments of the present application, on the one hand, the desensitization structured data set is obtained periodically and preprocessed to ensure the timeliness and accuracy of the information. Meanwhile, the preset portrait label library provides standardized labels that are easy for non-professional service personnel to understand. The portrait labels that match the treatment stage parameters and the emotional quantification parameters can be matched from these labels, which greatly simplifies the information presentation method, not only enables non-professional service personnel to quickly obtain key information, improves the service efficiency, but also improves the communication quality with patients and improves the patient satisfaction. On the other hand, the desensitization structured data set not only protects the privacy of patients and reduces the risk of privacy leakage, but also ensures the compliance of data processing and enhances the trust of patients in medical services.

[0020] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0021] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0022] Figure 1 is a method flow diagram of a reproductive diagnosis and treatment patient portrait generation method for non-professional service personnel provided by an embodiment of the present application; Figure 2 is a UI interface diagram related to a preset reproductive diagnosis and treatment knowledge graph provided by an embodiment of the present application; Figure 3 is a UI interface diagram related to a preset portrait label library provided by an embodiment of the present application; Figure 4 is a UI interface diagram of a patient portrait displayed by a client of a non-professional service personnel provided by an embodiment of the present application; Figure 5 is an interaction diagram of a server and a client provided by an embodiment of the present application; Figure 6 is a method flow diagram of a training method of a mental state recognition model provided by an embodiment of the present application; Figure 7 is a structural diagram of a reproductive diagnosis and treatment patient portrait generation device for non-professional service personnel provided by an embodiment of the present application; Figure 8 is a structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0023] The following description and drawings sufficiently illustrate specific embodiments of the present application to enable one skilled in the art to practice them.

[0024] It should be noted that the described embodiments are merely some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0025] The following description refers to the accompanying drawings. Unless otherwise noted, like elements in different drawings represent the same or similar elements. The following description of the example embodiments is not meant to represent all embodiments in accordance with the present application. Rather, it is merely intended to represent some aspects of the devices and methods in accordance with the present application as detailed in the attached claims.

[0026] In the description of the present application, it should be understood that the terms "first", "second" and the like are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances. In addition, in the description of the present application, unless otherwise specified, "multiple" means two or more. The association between the associated objects is described as "and / or", which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. The character " / " generally represents that the associated objects before and after are a kind of "or" relationship.

[0027] At present, the existing patient portrait system mainly relies on data in electronic medical record (EMR), laboratory information system (LIS) and hospital information system (HIS). These systems generate patient portraits by integrating patient medical records, test reports and appointment history and other information.

[0028] The applicant of the present application realizes that the existing patient portrait system mainly serves the clinical decision of expert doctors and relies on a large number of professional terms. These terms are difficult for non-professional service personnel (such as hospital customer service and nurses) to understand, and it is difficult for them to understand at a glance. Before communication, they must check the information or repeatedly ask the doctor, which makes them unable to quickly obtain key information when communicating with patients, affecting service efficiency.

[0029] To solve the above problems, the application provides a reproductive diagnosis and treatment patient portrait generation method and device for non-professional service personnel, to solve the problems existing in the above related technical problems. In the embodiments of the application, on the one hand, the desensitization structured data set is obtained periodically and preprocessed, ensuring the timeliness and accuracy of the information. At the same time, the preset portrait label library provides standardized labels that are easy for non-professional service personnel to understand, and the portrait labels that match the treatment stage parameters and emotional quantification parameters can be matched from these labels, greatly simplifying the information presentation method, not only enabling non-professional service personnel to quickly obtain key information, improving service efficiency, but also improving the communication quality with patients and improving patient satisfaction. On the other hand, the desensitization structured data set not only protects the privacy of patients and reduces the risk of privacy leakage, but also ensures the compliance of data processing and enhances the trust of patients in medical services. The following will be described in detail by exemplary embodiments.

[0030] The following will be described in detail by exemplary embodiments. Figure 1 The following will be described in detail by exemplary embodiments. Figure 6 The reproductive diagnosis and treatment patient portrait generation method for non-professional service personnel provided in the embodiments of the application will be described in detail. The method can be realized by relying on a computer program and can run on a reproductive diagnosis and treatment patient portrait generation device for non-professional service personnel based on the von Neumann system. The computer program can be integrated in an application or run as an independent tool application.

[0031] Please refer to Figure 1 A flowchart of a reproductive diagnosis and treatment patient portrait generation method for non-professional service personnel provided in the embodiments of the application is provided, which is applied to a server. As shown in the figure, the method of the embodiments of the application includes the following steps: Figure 1 S101, periodically obtaining and preprocessing the original medical data of each reproductive diagnosis and treatment patient through the data interface provided by the HIS system / LIS system, to obtain the desensitization structured data set of each reproductive diagnosis and treatment patient; The HIS system is a comprehensive system for managing various medical information within a hospital, including basic information of patients, medical records, test results, treatment records, etc. The LIS system is a system specially used for managing laboratory test data in a hospital, including sample collection, test results, quality control, etc. The data interface is a channel for data interaction between systems, allowing one system to obtain or send data from another system. The desensitization structured data set is a data collection after desensitization and formatting. Desensitization processing refers to processing sensitive information in the data so that it cannot identify specific individuals, while retaining the usability of the data.

[0032] ​In some embodiments of the present application, the specific process of obtaining and preprocessing the original medical data of each reproductive treatment patient to obtain the desensitized structured data set of each reproductive treatment patient includes: obtaining the original medical data of each reproductive treatment patient; extracting a plurality of key field information related to reproductive treatment from the original medical data; calling a TF-IDF sensitive word recognition model and loading a reproductive medical exclusive dictionary; desensitizing the plurality of key field information according to the TF-IDF sensitive word recognition model and the reproductive medical exclusive dictionary to obtain the desensitized structured data set of each reproductive treatment patient.

[0033] Among them, the key field information is specific information field directly related to reproductive treatment extracted from the original medical data, such as the age, gender, current treatment stage, examination result, etc. of the patient. TF-IDF (Term Frequency-Inverse Document Frequency) is a commonly used weighting technique for information retrieval and text mining. In the desensitization process, the TF-IDF model is used to identify and mark sensitive words. The reproductive medical exclusive dictionary is a glossary specifically for the reproductive medical field, which contains sensitive words and professional terms that may appear in medical data.

[0034] In a possible implementation, the complete medical records of the patient are obtained from the HIS system or LIS system of the hospital regularly (for example, every day). For example, the following original medical data of the patient is obtained from the HIS system. The fields related to reproductive treatment are extracted from the original medical data, such as age, gender, current treatment stage, psychological state, etc. The TF-IDF model and the reproductive medical exclusive dictionary are loaded for identifying sensitive words. The sensitive words are, for example, the patient's name, ID number, case. The desensitization process is performed on the extracted key field information using the TF-IDF model and the reproductive medical exclusive dictionary, and the desensitized structured data set of each reproductive treatment patient can be obtained.

[0035] S102, according to the desensitized structured data set, constructing the treatment stage parameter and the emotion quantization parameter of each reproductive treatment patient; In some embodiments of the present application, the specific process of constructing the treatment stage parameter and the emotion quantization parameter of each reproductive treatment patient according to the desensitized structured data set includes: determining the current treatment stage of each reproductive treatment patient according to the treatment path node defined in the preset reproductive treatment knowledge graph and the desensitized structured data set; predicting the remaining treatment days of each reproductive treatment patient according to the desensitized structured data set; taking the current treatment stage and the remaining treatment days as the treatment stage parameter of each reproductive treatment patient; analyzing the communication preference and the psychological state of each reproductive treatment patient according to the desensitized structured data set as the emotion quantization parameter of each reproductive treatment patient.

[0036] The preset reproductive diagnosis and treatment knowledge graph is a predefined knowledge structure, which describes various stages in the reproductive diagnosis and treatment process and their mutual relationships. The current treatment stage is the reproductive diagnosis and treatment stage in which the patient is currently located. The remaining treatment days are the number of days required for the patient to complete the remaining treatment stage.

[0037] The treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph carries a plurality of clinical attributes, and each treatment path node is sequentially connected according to the order of the ovulation induction stage, the oocyte retrieval stage, the embryo culture stage, the embryo transfer stage, and the follow-up stage.

[0038] In some embodiments of the present application, the specific process of determining the current treatment stage of each reproductive diagnosis and treatment patient according to the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph and the desensitized structured data set includes: obtaining the clinical feature attributes and conditions carried by each treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph; establishing first structured data between each treatment path node and the clinical feature attributes and conditions carried by each treatment path node; extracting a plurality of field attributes related to each treatment path node from the desensitized structured data set as a plurality of actual field attributes of each treatment path node; establishing second structured data between each treatment path node and the plurality of actual field attributes of each treatment path node; traversing each target node in the order of the preset reproductive diagnosis and treatment knowledge graph to combine a target node sequence; and determining the current treatment stage of each reproductive diagnosis and treatment patient according to the target node sequence, the first structured data, and the second structured data.

[0039] The specific features and conditions carried by each treatment path node for determining whether the patient is in the stage are used to match the actual medical data of the patient to determine the current treatment stage of the patient. The target node sequence is a sequence of treatment path nodes arranged in the order of the preset reproductive diagnosis and treatment knowledge graph.

[0040] The position index of each target node in the target node sequence is unique.

[0041] Specifically, the specific process of determining the current treatment stage of each reproductive diagnosis and treatment patient according to the target node sequence, the first structured data and the second structured data comprises: for each target node in the target node sequence, obtaining the clinical feature attribute and the condition from the first structured data; for each target node in the target node sequence, obtaining the plurality of actual field attributes from the second structured data; matching the clinical feature attribute and the condition of each target node with the plurality of actual field attributes to determine whether the clinical feature attribute and the condition exist in the plurality of actual field attributes to obtain the determination result of each target node in the target node sequence; determining the last target node in the target node sequence whose determination result indicates that the clinical feature attribute and the condition exist as the current treatment stage of each reproductive diagnosis and treatment patient.

[0042] In a possible implementation, the target node sequence is traversed, and the clinical feature attribute and the condition are matched. For example, in the ovulation induction stage, the clinical feature attribute is ["start date", "end date", "drug use"], and the condition is ["drug use record exists"]. The actual field attribute is that the start date is 2025-08-01, the end date is 2025-08-10, and the drug use is yes. The determination result is that the condition is met. If there is no subsequent treatment stage that meets the condition, it is indicated that the current treatment stage is the ovulation induction stage.

[0043] In some embodiments of the present application, the specific process of generating the preset reproductive diagnosis and treatment knowledge graph comprises: reading the reproductive specialist medical record template and the clinical guideline; defining the node path of the ovulation induction stage, the oocyte retrieval stage, the embryo culture stage, the embryo transfer stage and the follow-up stage as the graph main node path through the reproductive specialist medical record template and the clinical guideline; traversing the historical anonymous case to count the clinical feature attribute and the condition related to each graph main node path; and labeling the clinical feature attribute and the condition corresponding to each graph main node path on each graph main node path to obtain the preset reproductive diagnosis and treatment knowledge graph. Through the background management system, the preset reproductive diagnosis and treatment knowledge graph can be viewed. The content displayed in the background of the preset reproductive diagnosis and treatment knowledge graph is shown in the following table. Figure 2

[0044] The reproductive specialist medical record template is specially used for the medical record of the reproductive specialist, and contains the basic information, medical history, examination result, treatment process and the like of the patient. The clinical guideline is a medical operation specification based on the best clinical practice and scientific research. The preset reproductive diagnosis and treatment knowledge graph is a pre-defined knowledge structure, which describes the relationship between each stage in the reproductive diagnosis and treatment process and each stage, as well as the clinical feature attribute and the condition of each stage.

[0045] ​In a possible implementation, a reproductive specialist medical record template is read from a hospital medical record management system, and relevant clinical guidelines are consulted. According to the reproductive specialist medical record template and the clinical guidelines, a path of each stage in the reproductive diagnosis and treatment process is defined. Historical anonymous cases are traversed, and clinical feature attributes and conditions related to each path of the main node of the graph are counted. The counted clinical feature attributes and conditions are marked on each path of the main node of the graph. The data structure of each stage is as follows: { "Ovulation induction stage": { "attributes": ["start date", "end date", "drug use"], "conditions": ["drug use record exists"] }, "Oocyte retrieval stage": { "attributes": ["oocyte retrieval date", "oocyte quantity"], "conditions": ["oocyte retrieval date is not empty"] }, "Embryo culture stage": { "attributes": ["embryo culture start date", "embryo quantity"], "conditions": ["embryo culture start date is not empty"] }, "Embryo transfer stage": { "attributes": ["transfer date", "transfer embryo quantity"], "conditions": ["transfer date is not empty"] }, "Follow-up stage": { "attributes": ["first follow-up date", "follow-up result"], "conditions": ["first follow-up date is not empty"] } }。

[0046] In some embodiments of the present application, the specific process of predicting the remaining treatment days for each reproductive diagnosis and treatment patient based on the desensitized structured data set includes: determining multiple historical completed treatment stages before the current treatment stage; obtaining the start date and end date of each historical completed treatment stage from the desensitized structured data set; performing difference calculation on the start date and end date of each historical completed treatment stage to obtain the historical treatment cycle of each historical completed treatment stage; calculating the average value and standard deviation of the historical completed treatment stage based on the historical treatment cycle of each historical completed treatment stage; calculating the remaining treatment days for each reproductive diagnosis and treatment patient based on the average time and standard deviation; the calculation formula for the remaining treatment days for each reproductive diagnosis and treatment patient is:

[0047] in, is the adjustment coefficient, which is adjusted according to clinical experience and data distribution. is the average time, is the standard deviation; Average time The calculation formula is:

[0048] in, is the number of historically completed treatment phases, It is historical treatment cycles with completed treatment phases; The formula for calculating the standard deviation is:

[0049] in, is the standard deviation.

[0050] In some embodiments of the present application, the specific process of analyzing the communication preferences and psychological state of each reproductive diagnosis and treatment patient based on the desensitized structured data set includes: extracting the communication records and psychological state-related field attributes of each reproductive diagnosis and treatment patient from the desensitized structured data set; counting the frequency of occurrence of each preset communication method of each reproductive diagnosis and treatment patient based on the communication records; taking the preset communication method with the highest frequency as the communication preference of each reproductive diagnosis and treatment patient; converting the psychological state-related field attributes into emotion description text to obtain the emotion description text of each reproductive diagnosis and treatment patient; inputting the emotion description text of each reproductive diagnosis and treatment patient into a pre-trained psychological state recognition model to output the psychological state of each reproductive diagnosis and treatment patient.

[0051] Specifically, the specific process of generating the pre-trained psychological state recognition model includes: collecting a plurality of historical emotional description texts of each historical reproductive diagnosis and treatment patient; labeling a psychological state label for each historical emotional description text to obtain a model training sample; constructing a psychological state recognition model using a neural network, and performing machine learning on the psychological state recognition model using the model training sample to obtain the pre-trained psychological state recognition model.

[0052] For example, for a patient with a patient ID of A12345, the final treatment stage parameter is: the current treatment stage is the ovulation stage, and the remaining treatment days are 7 days; the emotional quantification parameter is: the communication preference is telephone; and the psychological state is anxiety.

[0053] S103, from the preset portrait label library, match a plurality of target portrait labels belonging to the treatment stage parameter and the emotional quantification parameter; The treatment stage parameter includes the current treatment stage and the remaining treatment days; the emotional quantification parameter includes the communication preference and the psychological state; and the preset portrait label library includes each portrait label and the portrait parameter condition corresponding to each portrait label, and the portrait parameter condition includes the treatment stage condition and the emotional quantification condition. Through the background management system, the preset portrait label library can be viewed, and the content displayed in the background of the preset portrait label library is, for example Figure 3 as shown.

[0054] In some embodiments of the present application, the specific process of matching a plurality of target portrait labels belonging to the treatment stage parameter and the emotional quantification parameter from the preset portrait label library includes: traversing each target portrait label in the preset portrait label library; analyzing whether the current treatment stage and the remaining treatment days satisfy the treatment stage condition corresponding to each traversed target portrait label; analyzing whether the communication preference and the psychological state satisfy the emotional quantification condition corresponding to each traversed target portrait label; and caching the target portrait label satisfying the treatment stage condition and the emotional quantification condition as the plurality of target portrait labels belonging to the treatment stage parameter and the emotional quantification parameter.

[0055] The treatment stage condition is a condition related to the treatment stage defined in each target portrait label, such as the current treatment stage and the remaining treatment days. The emotional quantification condition is a condition related to the emotional state defined in each target portrait label, such as the communication preference and the psychological state.

[0056] In one possible implementation, for patient ID A12345, the final treatment stage parameters are: the current treatment stage is ovulation induction, and the remaining treatment days are 7 days; the emotional quantification parameters are: the communication preference is phone calls, and the psychological state is anxiety. The portrait label library contains two example labels. Label L001 describes the patient as being in the ovulation induction stage, with fewer than 10 days remaining, the communication preference is phone calls, and the psychological state is anxiety. The treatment stage conditions are: the current treatment stage is "ovulation induction," with fewer than 10 days remaining. The emotional quantification conditions are: the communication preference is "phone calls," and the psychological state is "anxiety." Label L002 describes the patient as being in the egg retrieval stage, with fewer than 5 days remaining, the communication preference is text messages, and the psychological state is optimistic. The treatment stage conditions are: the current treatment stage is "egg retrieval," with fewer than 5 days remaining. The emotional quantification conditions are: the communication preference is "text messages," and the psychological state is "optimism."

[0057] For example, analyzing the treatment stage condition, for label L001: Check whether the patient's current treatment stage is "ovulation induction": Yes, the patient is currently in ovulation induction. Check whether the remaining treatment days are less than 10: Yes, the patient has 7 days remaining, which is less than 10. Therefore, the patient meets the treatment stage condition for label L001. For label L002: Check whether the patient's current treatment stage is "egg retrieval": No, the patient is currently in ovulation induction. Therefore, the patient does not meet the treatment stage condition for label L002, and no further emotional quantification conditions are required. Analyzing the emotional quantification conditions, for label L001: Check whether the patient's communication preference is "telephone": Yes, the patient prefers telephone communication. Check whether the patient's psychological state is "anxiety": Yes, the patient's current psychological state is anxious. Therefore, the patient also meets the emotional quantification condition for label L001. After the above analysis, we find that the patient fully meets the description of label L001. Therefore, we cache label L001 as the patient's profile label.

[0058] S104, filling multiple target portrait tags into a preset natural language template to obtain a portrait description text of each reproductive diagnosis and treatment patient, and sending it to the client of non-professional service personnel for display.

[0059] The preset natural language template is a pre-designed text template used to convert profile labels into easy-to-understand natural language descriptions. The client for non-professional service personnel is a device or software interface used by non-professional service personnel (such as customer service representatives and nurses) to display patient profile description text.

[0060] In some embodiments of the present application, the target patient ID12345 has the following two target portrait labels. Label L001 describes that the patient is currently in the ovulation induction stage, the expected remaining treatment days are 7 days, the preferred communication method is through the phone, and the psychological state is anxiety. Label L003 describes that the patient has high expectations for the treatment process and hopes to receive more attention and support.

[0061] For example, the natural language template is: The patient {patient_id} is currently in {treatment_stage}, and the expected remaining treatment days are {remaining_days} days. The patient prefers to communicate through {communication_preference}, and the current psychological state is {psychological_state}. In addition, {additional_info}.

[0062] Finally, according to the target portrait label, the multiple target portrait labels are filled into the preset natural language template to obtain the filled portrait description text. For example: The patient A12345 is currently in the ovulation induction stage, and the expected remaining treatment days are 7 days. The patient prefers to communicate through the phone, and the current psychological state is anxiety. In addition, the patient has high expectations for the treatment process and hopes to receive more attention and support. The result displayed on the client of the non-professional service personnel is shown as follows: Figure 4

[0063] For example Figure 5 As shown, the server 110 regularly obtains and pre-processes the original medical data of each reproductive diagnosis and treatment patient through the data interface provided by the HIS system / LIS system to obtain the desensitized structured data set of each reproductive diagnosis and treatment patient. The server 110 constructs the treatment stage parameter and the emotion quantization parameter of each reproductive diagnosis and treatment patient according to the desensitized structured data set. The server 110 matches multiple target portrait labels belonging to the treatment stage parameter and the emotion quantization parameter from the preset portrait label library. The server 110 fills the multiple target portrait labels into the preset natural language template to obtain the portrait description text of each reproductive diagnosis and treatment patient, and sends it to the client 120 of the non-professional service personnel for display.

[0064] ​In the embodiments of the present application, on the one hand, the desensitization structured data set is obtained periodically and preprocessed, ensuring the timeliness and accuracy of the information. Meanwhile, the preset portrait label library provides standardized labels that are easy for non-professional service personnel to understand. The portrait labels that match the treatment stage parameters and the emotional quantification parameters can be matched from these labels, greatly simplifying the information presentation method, not only enabling non-professional service personnel to quickly obtain key information and improving service efficiency, but also improving the communication quality with patients and improving patient satisfaction. On the other hand, the desensitization structured data set not only protects the privacy of patients and reduces the risk of privacy leakage, but also ensures the compliance of data processing and enhances the trust of patients in medical services.

[0065] Please refer to Figure 6 , a flowchart of a large language model fine-tuning method is provided for the embodiments of the present application. As shown in Figure 6 , the method of the embodiments of the present application can include the following steps: S201, collecting a plurality of historical emotional description texts of each historical reproductive diagnosis and treatment patient; S202, labeling a psychological state label for each historical emotional description text to obtain a model training sample; S203, constructing a psychological state recognition model using a neural network, and performing machine learning on the psychological state recognition model using the model training sample to obtain a pre-trained psychological state recognition model.

[0066] In the embodiments of the present application, by collecting a large number of historical emotional description texts and labeling psychological state labels, the model can learn the complex mapping relationship between different emotional descriptions and psychological states. The trained model can more accurately identify the psychological state of the patient in actual application, reducing misjudgment and omission, thereby providing more accurate psychological support for the patient.

[0067] The following is an apparatus embodiment of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the apparatus embodiments of the present application, please refer to the method embodiments of the present application.

[0068] Please refer to Figure 7 , which shows a structural schematic diagram of a reproductive diagnosis and treatment patient portrait generation device for non-professional service personnel provided by an exemplary embodiment of the present application. The reproductive diagnosis and treatment patient portrait generation device for non-professional service personnel can be realized as all or part of an electronic device through software, hardware or a combination of the two. The device 1 includes a data preprocessing module 10, a parameter construction module 20, a portrait label matching module 30, and a portrait description text display module 40.

[0069] The data preprocessing module 10 is configured to periodically acquire and preprocess original medical data of each reproductive diagnosis and treatment patient through a data interface provided by a HIS system / LIS system to obtain a desensitized structured data set of each reproductive diagnosis and treatment patient. The parameter construction module 20 is configured to construct a treatment stage parameter and an emotion quantization parameter of each reproductive diagnosis and treatment patient according to the desensitized structured data set. The image label matching module 30 is configured to match a plurality of target image labels belonging to the treatment stage parameter and the emotion quantization parameter from a preset image label library. The image description text display module 40 is configured to fill the plurality of target image labels into a preset natural language template to obtain an image description text of each reproductive diagnosis and treatment patient, and send the image description text to a client of a non-professional service personnel for display.

[0070] It should be noted that the reproductive diagnosis and treatment patient image generation device for non-professional service personnel provided in the above embodiments is only exemplified by the division of the above functional modules when performing the reproductive diagnosis and treatment patient image generation method for non-professional service personnel. In actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the reproductive diagnosis and treatment patient image generation device for non-professional service personnel and the reproductive diagnosis and treatment patient image generation method for non-professional service personnel provided in the above embodiments belong to the same concept, and the implementation process is detailed in the method embodiments. Here, it is not repeated.

[0071] The serial numbers of the above embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0072] In the embodiments of the present application, on the one hand, the desensitized structured data set is periodically acquired and preprocessed to ensure the timeliness and accuracy of the information. At the same time, the preset image label library provides standardized labels that are easy for non-professional service personnel to understand. The image labels matched from these labels are matched with the treatment stage parameter and the emotion quantization parameter, which greatly simplifies the information presentation method, not only enables non-professional service personnel to quickly obtain key information, improves service efficiency, but also improves the communication quality with patients and improves patient satisfaction. On the other hand, the desensitized structured data set not only protects the privacy of patients and reduces the risk of privacy leakage, but also ensures the compliance of data processing and enhances the trust of patients in medical services.

[0073] The present application also provides a computer readable medium having program instructions stored thereon, which, when executed by a processor, implement the reproductive diagnosis and treatment patient image generation method for non-professional service personnel provided by each of the above method embodiments.

[0074] The application further provides a computer program product comprising instructions which, when executed on a computer, cause the computer to perform the non-professional service personnel-oriented reproductive diagnosis and treatment patient portrait generation method of each method embodiment described above.

[0075] See Figure 8 A structural schematic diagram of an electronic device is provided for the embodiments of the application. As shown in the figure, Figure 8 The electronic device 1000 can include at least one processor 1001, at least one network interface 1004, a user interface 1003, a memory 1005, and at least one communication bus 1002.

[0076] The communication bus 1002 is used to realize the connection and communication between the components.

[0077] The user interface 1003 can include a display screen (Display) and a camera (Camera). Optionally, the user interface 1003 can further include a standard wired interface and a wireless interface.

[0078] The network interface 1004 can optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0079] The processor 1001 can include one or more processing cores. The processor 1001 connects various parts of the entire electronic device 1000 through various interfaces and lines, executes various functions of the electronic device 1000 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 1005, and calling data stored in the memory 1005. Optionally, the processor 1001 can be realized in at least one of the hardware forms of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 1001 can integrate a combination of one or several of central processing units (CPUs), graphics processing units (GPUs), and modems. Among them, the CPU is mainly used to process operating systems, user interfaces, and application programs; the GPU is used to render and draw the content to be displayed on the display screen; and the modem is used to process wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 1001, but can be realized by a separate chip.

[0080] The memory 1005 can include a random access memory (RAM) and can also include a read-only memory (ROM). Optionally, the memory 1005 includes a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, codes, code sets, or instruction sets. The memory 1005 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the various method embodiments described above, etc.; the data storage area can store data involved in the various method embodiments described above, etc. The memory 1005 can also be at least one storage system located away from the aforementioned processor 1001. As shown in Figure 8 The memory 1005 as a computer storage medium can include an operating system, a network communication module, a user interface module, and a patient portrait generation application for non-professional service personnel in reproductive diagnosis and treatment.

[0081] In Figure 8 In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an interface for user input and obtain user input data; and the processor 1001 can be used to call the patient portrait generation application for non-professional service personnel in reproductive diagnosis and treatment stored in the memory 1005 and specifically perform the following operations: Obtain and preprocess the original medical data of each patient in reproductive diagnosis and treatment through the data interface provided by the HIS system / LIS system to obtain the desensitized structured data set of each patient in reproductive diagnosis and treatment; According to the desensitized structured data set, construct the treatment stage parameter and the emotion quantization parameter of each patient in reproductive diagnosis and treatment; Match a plurality of target portrait labels belonging to the treatment stage parameter and the emotion quantization parameter from the preset portrait label library; Fill the plurality of target portrait labels into the preset natural language template to obtain the portrait description text of each patient in reproductive diagnosis and treatment, and send it to the client of the non-professional service personnel for display.

[0082] In one embodiment, when the processor 1001 executes the operation of constructing the treatment stage parameter and the emotion quantization parameter of each patient in reproductive diagnosis and treatment according to the desensitized structured data set, it specifically performs the following operations: Determine the current treatment stage of each patient in reproductive diagnosis and treatment according to the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph and the desensitized structured data set; According to the desensitization structured data set, the remaining treatment days of each reproductive treatment patient are predicted; The current treatment stage and the remaining treatment days are taken as the treatment stage parameters of each reproductive treatment patient. According to the desensitization structured data set, the communication preferences and psychological states of each reproductive treatment patient are analyzed as the emotional quantification parameters of each reproductive treatment patient.

[0083] In one embodiment, when the processor 1001 determines the current treatment stage of each reproductive treatment patient according to the treatment path nodes defined in the preset reproductive treatment knowledge graph and the desensitization structured data set, the following operations are specifically performed: From the preset reproductive treatment knowledge graph, the clinical feature attributes and conditions carried by each defined treatment path node are obtained. First structured data is established between each treatment path node and the clinical feature attributes and conditions carried by each treatment path node. From the desensitization structured data set, a plurality of field attributes related to each treatment path node are extracted as a plurality of actual field attributes of each treatment path node. Second structured data is established between each treatment path node and the plurality of actual field attributes of each treatment path node. According to the preset reproductive treatment knowledge graph, each target node is traversed in sequence to form a target node sequence. According to the target node sequence, the first structured data, and the second structured data, the current treatment stage of each reproductive treatment patient is determined.

[0084] In one embodiment, when the processor 1001 determines the current treatment stage of each reproductive treatment patient according to the target node sequence, the first structured data, and the second structured data, the following operations are specifically performed: For each target node in the target node sequence, the clinical feature attributes and conditions are obtained from the first structured data. For each target node in the target node sequence, the plurality of actual field attributes are obtained from the second structured data. The clinical feature attributes and conditions of each target node are matched with the plurality of actual field attributes to determine whether the clinical feature attributes and conditions exist in the plurality of actual field attributes, and a judgment result of each target node in the target node sequence is obtained. The last target node in the target node sequence whose judgment result indicates the existence of the clinical feature attributes and conditions is determined as the current treatment stage of each reproductive treatment patient.

[0085] In one embodiment, when the processor 1001 generates the preset reproductive treatment knowledge graph, the following operations are specifically performed: reading the reproductive specialist medical record template and the clinical guideline; Defining the node path of the ovulation promotion stage, the oocyte retrieval stage, the embryo culture stage, the embryo transfer stage, and the follow-up stage as the main node path of the atlas through the reproductive specialist medical record template and the clinical guideline; Traversing the historical anonymous cases to count the clinical feature attributes and conditions related to each main node path of the atlas; Labeling the clinical feature attributes and conditions corresponding to each main node path of the atlas on each main node path of the atlas to obtain the preset reproductive diagnosis and treatment knowledge graph.

[0086] In one embodiment, the processor 1001 specifically performs the following operations when predicting the remaining treatment days of each reproductive diagnosis and treatment patient according to the de-structured structured data set: Determine a plurality of historical completed treatment stages before the current treatment stage; Obtain the start date and end date of each historical completed treatment stage from the de-structured structured data set; Calculate the difference between the start date and the end date of each historical completed treatment stage to obtain the historical treatment period of each historical completed treatment stage; Calculate the mean and standard deviation of the historical completed treatment stage according to the historical treatment period of each historical completed treatment stage; Calculate the remaining treatment days of each reproductive diagnosis and treatment patient according to the average time consumption and the standard deviation; The calculation formula of the remaining treatment days of each reproductive diagnosis and treatment patient is:

[0087] wherein, is an adjustment coefficient, which is adjusted according to clinical experience and data distribution, is the average time consumption, is the standard deviation; The calculation formula of the average time consumption is:

[0088] wherein, is the number of historical completed treatment stages, is the historical treatment period of the th historical completed treatment stage; The calculation formula of the standard deviation is:

[0089] wherein, is the standard deviation.

[0090] In one embodiment, the processor 1001 specifically performs the following operations when performing analysis on the communication preferences and psychological states of each reproductive treatment patient according to the desensitized structured data set: Extracting the communication records and psychological state related field attributes of each reproductive treatment patient from the desensitized structured data set; According to the communication records, the occurrence frequency of each pre-set communication mode of each reproductive treatment patient is counted; The pre-set communication mode with the highest occurrence frequency is taken as the communication preference of each reproductive treatment patient; The psychological state related field attributes are converted into emotional description text to obtain the emotional description text of each reproductive treatment patient; The emotional description text of each reproductive treatment patient is input into a pre-trained psychological state recognition model to output the psychological state of each reproductive treatment patient.

[0091] In one embodiment, the processor 1001 specifically performs the following operations when performing matching of multiple target portrait labels belonging to the treatment stage parameters and emotional quantification parameters from the pre-set portrait label library: Traverse each target portrait label in the pre-set portrait label library; Analyze whether the current treatment stage and the remaining treatment days meet the treatment stage condition corresponding to each target portrait label in the traversal; Analyze whether the communication preference and the psychological state meet the emotional quantification condition corresponding to each target portrait label in the traversal; Cache the target portrait labels that meet the treatment stage condition and the emotional quantification condition as the multiple target portrait labels belonging to the treatment stage parameters and the emotional quantification parameters.

[0092] In one embodiment, the processor 1001 specifically performs the following operations when performing acquisition and preprocessing of the original medical data of each reproductive treatment patient to obtain the desensitized structured data set of each reproductive treatment patient: Acquire the original medical data of each reproductive treatment patient; Extracting a plurality of key field information related to reproductive treatment from the original medical data; Calling a TF-IDF sensitive word recognition model and loading a reproductive medical exclusive dictionary; According to the TF-IDF sensitive word recognition model and the reproductive medical exclusive dictionary, desensitizing the plurality of key field information to obtain the desensitized structured data set of each reproductive treatment patient.

[0093] In the embodiments of the present application, on the one hand, the desensitization structured data set is obtained periodically and preprocessed, ensuring the timeliness and accuracy of the information. Meanwhile, the preset portrait label library provides standardized labels that are easy for non-professional service personnel to understand. The portrait labels that match the treatment stage parameters and the emotional quantification parameters can be matched from these labels, greatly simplifying the information presentation method, not only enabling non-professional service personnel to quickly obtain key information and improving service efficiency, but also improving the communication quality with patients and improving patient satisfaction. On the other hand, the desensitization structured data set not only protects the privacy of patients and reduces the risk of privacy leakage, but also ensures the compliance of data processing and enhances the trust of patients in medical services.

[0094] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware. The program for generating a portrait of a patient for reproductive diagnosis and treatment facing non-professional service personnel can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments of each method. The storage medium of the program for generating a portrait of a patient for reproductive diagnosis and treatment facing non-professional service personnel can be a disk, an optical disk, a read-only memory or a random access memory, etc.

[0095] The above only discloses the preferred embodiments of the present application, and of course cannot limit the scope of the rights of the present application. Therefore, equivalent changes made in accordance with the claims of the present application are still within the scope covered by the present application.

Claims

1. A method for generating a portrait of a patient for reproductive treatment for a non-professional service personnel, characterized in that, Applied to a server, the method comprises: Periodically acquiring and preprocessing original medical data of each reproductive diagnosis and treatment patient through a data interface provided by a HIS system / LIS system to obtain a desensitized structured data set of each reproductive diagnosis and treatment patient; According to the desensitized structured data set, a treatment stage parameter and an emotional quantization parameter of each reproductive diagnosis and treatment patient are constructed; From a preset portrait label library, a plurality of target portrait labels belonging to the treatment stage parameter and the emotional quantization parameter are matched; The plurality of target portrait labels are filled into a preset natural language template to obtain a portrait description text of each reproductive diagnosis and treatment patient, which is sent to a client of a non-professional service personnel for display.

2. The method of claim 1, wherein, According to the desensitized structured data set, the treatment stage parameter and the emotional quantization parameter of each reproductive diagnosis and treatment patient are constructed, comprising: According to a treatment path node defined in a preset reproductive diagnosis and treatment knowledge graph and the desensitized structured data set, a current treatment stage of each reproductive diagnosis and treatment patient is determined; According to the desensitized structured data set, a remaining treatment day of each reproductive diagnosis and treatment patient is predicted; The current treatment stage and the remaining treatment day are taken as the treatment stage parameter of each reproductive diagnosis and treatment patient; According to the desensitized structured data set, a communication preference and a psychological state of each reproductive diagnosis and treatment patient are analyzed as the emotional quantization parameter of each reproductive diagnosis and treatment patient.

3. The method of claim 2, wherein, The treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph carries a plurality of clinical attributes, and each treatment path node is sequentially connected in the order of ovulation induction stage, ovum collection stage, embryo culture stage, embryo transfer stage and follow-up stage; According to the treatment path node defined in the preset reproductive diagnosis and treatment knowledge graph and the desensitized structured data set, the current treatment stage of each reproductive diagnosis and treatment patient is determined, comprising: From the preset reproductive diagnosis and treatment knowledge graph, clinical feature attributes and conditions carried by each defined treatment path node are acquired; First structured data between each treatment path node and the clinical feature attributes and conditions carried by each treatment path node are established; From the desensitized structured data set, a plurality of field attributes related to each treatment path node are extracted as a plurality of actual field attributes of each treatment path node; Second structured data between each treatment path node and the plurality of actual field attributes of each treatment path node are established; Each target node is traversed in the order of the preset reproductive diagnosis and treatment knowledge graph to be combined into a target node sequence; According to the target node sequence, the first structured data and the second structured data, the current treatment stage of each reproductive diagnosis and treatment patient is determined.

4. The method of claim 3, wherein, The position index of each target node in the target node sequence is unique. According to the target node sequence, the first structured data and the second structured data, the current treatment stage of each reproductive diagnosis and treatment patient is determined, comprising: For each target node in the target node sequence, clinical feature attributes and conditions are acquired from the first structured data; For each target node in the target node sequence, obtain a plurality of actual field attributes from the second structured data; Match the clinical feature attributes and conditions of each target node with the plurality of actual field attributes to determine whether the clinical feature attributes and conditions exist in the plurality of actual field attributes, and obtain a judgment result of each target node in the target node sequence; Determine the target node indicated by the last judgment result in the target node sequence as the current treatment stage of each reproductive diagnosis and treatment patient.

5. The method of claim 2, wherein, The preset reproductive diagnosis and treatment knowledge graph is generated by the following steps, including: reading reproductive specialist medical record templates and clinical guidelines; Defining the node paths of the ovulation induction stage, oocyte retrieval stage, embryo culture stage, embryo transfer stage, and follow-up stage as the main node paths of the graph through the reproductive specialist medical record templates and clinical guidelines; Traverse the historical anonymous cases to count the clinical feature attributes and conditions related to each main node path of the graph; Label the clinical feature attributes and conditions corresponding to each main node path of the graph on each main node path of the graph to obtain the preset reproductive diagnosis and treatment knowledge graph.

6. The method of claim 2, wherein, The remaining treatment days of each reproductive diagnosis and treatment patient are predicted according to the desensitized structured data set, including: Determine a plurality of historical completed treatment stages before the current treatment stage; Obtain the start date and end date of each historical completed treatment stage from the desensitized structured data set; Calculate the difference between the start date and the end date of each historical completed treatment stage to obtain the historical treatment cycle of each historical completed treatment stage; Calculate the average value and standard deviation of the historical completed treatment stages according to the historical treatment cycle of each historical completed treatment stage; According to the average time consumption and standard deviation, the remaining treatment days of each reproductive diagnosis and treatment patient are calculated; The calculation formula of the remaining treatment days of each reproductive diagnosis and treatment patient is: wherein, is an adjustment factor, which is adjusted according to clinical experience and data distribution, is the average time consumption, is the standard deviation; The average time consumption The calculation formula is: wherein, is the number of historical completed treatment phases, is the historical treatment cycle of the th historical completed treatment phase; The calculation formula of the standard deviation is: wherein is the standard deviation.

7. The method of claim 2, wherein, According to the desensitized structured data set, the communication preference and psychological state of each reproductive diagnosis and treatment patient are analyzed, including: Extract the communication records and psychological state related field attributes of each reproductive diagnosis and treatment patient from the desensitized structured data set; According to the communication records, count the occurrence frequency of each preset communication mode of each reproductive diagnosis and treatment patient; The preset communication mode with the highest occurrence frequency is taken as the communication preference of each reproductive diagnosis and treatment patient; Convert the psychological state related field attributes into emotion description text to obtain the emotion description text of each reproductive diagnosis and treatment patient; Input the emotion description text of each reproductive diagnosis and treatment patient into a pre-trained psychological state recognition model to output the psychological state of each reproductive diagnosis and treatment patient; wherein The pre-trained psychological state recognition model is generated by the following steps, including: Collecting a plurality of historical emotion description texts of each historical reproductive diagnosis and treatment patient; Label the psychological state label of each historical emotion description text to obtain model training samples; The psychological state recognition model is constructed by using a neural network, and a sample is trained by using the model to perform machine learning on the psychological state recognition model, so as to obtain a pre-trained psychological state recognition model.

8. The method according to any one of claims 1 to 7, characterized in that, The treatment stage parameter includes a current treatment stage and a remaining treatment day number, and the emotion quantification parameter includes a communication preference and a psychological state; the preset portrait label library includes each portrait label and a portrait parameter condition corresponding to each portrait label, and the portrait parameter condition includes a treatment stage condition and an emotion quantification condition; The matching of the multiple target portrait labels belonging to the treatment stage parameter and the emotion quantification parameter from the preset portrait label library includes: Traversing each target portrait label in the preset portrait label library; Analyzing whether the current treatment stage and the remaining treatment day number satisfy the treatment stage condition corresponding to each traversed target portrait label; Analyzing whether the communication preference and the psychological state satisfy the emotion quantification condition corresponding to each traversed target portrait label; Caching the target portrait label satisfying the treatment stage condition and the emotion quantification condition as the multiple target portrait labels belonging to the treatment stage parameter and the emotion quantification parameter.

9. The method according to any one of claims 1 to 7, characterized in that, The obtaining and preprocessing of the original medical data of each reproductive diagnosis and treatment patient to obtain the desensitization structured data set of each reproductive diagnosis and treatment patient includes: Obtaining the original medical data of each reproductive diagnosis and treatment patient; Extracting multiple key field information related to reproductive diagnosis and treatment from the original medical data; Calling a TF-IDF sensitive word recognition model and loading a reproductive medical exclusive dictionary; Desensitizing the multiple key field information according to the TF-IDF sensitive word recognition model and the reproductive medical exclusive dictionary to obtain the desensitization structured data set of each reproductive diagnosis and treatment patient.

10. A device for generating a portrait of a patient for a non-specialist service provider in a reproductive medical treatment, characterized by The device includes: A data preprocessing module configured to periodically obtain and preprocess original medical data of each reproductive diagnosis and treatment patient through a data interface provided by a HIS system / LIS system to obtain a desensitization structured data set of each reproductive diagnosis and treatment patient; A parameter construction module configured to construct a treatment stage parameter and an emotion quantification parameter of each reproductive diagnosis and treatment patient according to the desensitization structured data set; A portrait label matching module configured to match multiple target portrait labels belonging to the treatment stage parameter and the emotion quantification parameter from a preset portrait label library; A portrait description text display module configured to fill the multiple target portrait labels into a preset natural language template to obtain a portrait description text of each reproductive diagnosis and treatment patient and send the portrait description text to a client of a non-professional service personnel for display.

Images