An information determination method, apparatus, device, computer storage medium, and computer program product
By assessing the attributes and symptoms of emergency patients and using language processing and hierarchical models for condition evaluation, the accuracy and efficiency of vehicle allocation and guidance in emergency dispatch have been improved, resulting in more efficient allocation and guidance of emergency resources.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA MOBILE CHENGDU INFORMATION & TELECOMM TECH CO LTD
- Filing Date
- 2024-11-28
- Publication Date
- 2026-05-29
AI Technical Summary
In emergency dispatch, the accuracy and efficiency of vehicle allocation are poor, the accuracy of emergency guidance is low, and the complete reliance on the operator's professional knowledge and experience leads to patients not receiving timely treatment.
By identifying the target's attribute and symptom information, language processing and hierarchical models are used to assess the severity of the condition. This information is then combined with target guidance information to allocate vehicles and implement emergency measures, reducing reliance on the operator's specialized knowledge.
It improved the accuracy and efficiency of emergency vehicle allocation, ensured that patients received timely and accurate emergency guidance, and enhanced the overall efficiency and accuracy of emergency dispatch.
Smart Images
Figure CN122117276A_ABST
Abstract
Description
Technical Field
[0001] This application relates to information determination technology in the field of network technology, and more particularly to an information determination method, apparatus, device, computer storage medium, and computer program product. Background Technology
[0002] With the rapid development of emergency medicine, more and more emergency patients are receiving timely treatment. However, as is well known, the golden time for emergency care is extremely short. Therefore, during emergency dispatch, it is crucial to quickly gather relevant patient information and make a preliminary diagnosis of the patient's condition for appropriate emergency dispatch. Currently, related technologies typically involve operators manually entering patient information and making a preliminary diagnosis based on this information. This preliminary diagnosis then leads to the allocation of rescue vehicles. During the emergency dispatch process, personnel determine emergency measures based on the preliminary diagnosis and provide emergency care guidance to the patient before the ambulance arrives at the scene.
[0003] However, the relevant technologies rely entirely on the operator's professional knowledge when dispatching emergency medical services, which leads to poor accuracy and low efficiency in vehicle allocation, resulting in patients not receiving timely treatment. Furthermore, the reliance on the operator's experience to provide rescue guidance makes it impossible to guarantee the accuracy of emergency medical guidance. Summary of the Invention
[0004] To address the aforementioned technical problems, this application aims to provide an information determination method, apparatus, device, computer storage medium, and computer program product, which solves the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy in emergency dispatching in related technologies.
[0005] The technical solution of this application is implemented as follows:
[0006] An information determination method, the method comprising:
[0007] Determine the attribute information and target symptom information of the target object; wherein, the target symptom information is information related to the condition of the target object;
[0008] Based on the attribute information and the target symptom information, a target level for the condition and target guidance information for the target individual are determined; wherein, the target level characterizes the severity of the condition; and the target guidance information is used to alleviate the condition of the target individual.
[0009] Based on the target language processing model, the target level, and the target guidance information, target feedback information is determined for the target object.
[0010] In the above scheme, determining the attribute information and target symptom information of the target object includes:
[0011] Obtain the dialogue information corresponding to the target object;
[0012] The target language processing model is used to process the dialogue information to obtain the initial attribute information and initial symptom information of the target object;
[0013] The initial attribute information and the initial symptom information are validated to obtain the attribute information and the target symptom information.
[0014] In the above scheme, determining the target level of the condition based on the attribute information and the target symptom information includes:
[0015] Based on the attribute information and the target symptom information, the initial level of the condition is determined;
[0016] Obtain the target number and target parameters of the objects to be processed at the initial level; wherein, the target parameters characterize the usage status of the rescue vehicle;
[0017] The target symptom information, the number of targets, and the target parameters are processed using a target adjustment model to obtain the target level.
[0018] In the above scheme, determining the initial level of the condition based on the attribute information and the target symptom information includes:
[0019] Based on the attribute information and the target symptom information, the first category of the target object is determined;
[0020] If, based on the first category and the target mapping relationship, it is determined that there is matching symptom information in the target database that matches the target symptom information, the level corresponding to the matching symptom information is determined as the initial level; wherein, the target mapping relationship is the mapping relationship between disease category and symptom information;
[0021] If it is determined that the matching symptom information does not exist in the target database, the attribute information and the target symptom information are processed using a target grading model to obtain the initial level.
[0022] The method in the above scheme further includes:
[0023] Obtain historical attribute information and historical symptom information of the historical object; wherein, the historical symptom information is information related to the illness of the historical object;
[0024] Based on the historical attribute information and the historical symptom information, the historical level of the patient's condition is determined;
[0025] Based on historical data, historical parameters, and the historical symptom information, the initial adjustment model is trained to obtain the target adjustment model; wherein, the historical data refers to the number of historical objects at the historical level; and the historical parameters characterize the historical usage of the rescue vehicle.
[0026] In the above scheme, based on the attribute information and the target symptom information, target guidance information for the target object is determined, including:
[0027] The attribute information and the target symptom information are processed using a target classification model to obtain the target category of the disease.
[0028] If a second category matching the target category exists in the disease categories stored in the target database, the guidance information corresponding to the second category is determined as the target guidance information;
[0029] If the second category does not exist in the disease category, the target language processing model is used to process the target category, the attribute information, and the symptom information to obtain the target guidance information.
[0030] The method in the above scheme further includes:
[0031] Retrieve historical dialogue information corresponding to historical objects;
[0032] The target language processing model is used to process the historical dialogue information to obtain the historical attribute information and the historical symptom information;
[0033] Based on the historical attribute information and the historical symptom information, the initial classification model is trained to obtain the target classification model.
[0034] In the above scheme, determining the target feedback information for the target object based on the target language processing model, the target level, and the target guidance information includes:
[0035] The target language processing model is used to summarize and process the target level and the target guidance information to obtain initial feedback information for the target object;
[0036] The initial feedback information is converted to obtain the target feedback information;
[0037] The method in the above scheme further includes:
[0038] The target database is updated based on the dialogue information corresponding to the target object and the target feedback information.
[0039] An information determining device, the device comprising:
[0040] The first determining unit is used to determine the attribute information and target symptom information of the target object; wherein the target symptom information is information related to the condition of the target object;
[0041] The second determining unit is configured to determine the target level of the illness and target guidance information for the target object based on the attribute information and the target symptom information; wherein the target level characterizes the severity of the illness; and the target guidance information is used to alleviate the illness of the target object.
[0042] The third determining unit is used to determine target feedback information for the target object based on the target language processing model, the target level, and the target guidance information.
[0043] An information determining device, the device comprising: a processor, a memory, and a communication bus;
[0044] The communication bus is used to realize the communication connection between the processor and the memory;
[0045] The processor is used to execute the information determination program in the memory to implement the steps of the information determination method described above.
[0046] A computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the steps of the information determination method described above.
[0047] A computer program product comprising a computer program that, when executed by a processor, implements the aforementioned information determination method.
[0048] The information determination method, apparatus, device, computer storage medium, and computer program product provided in this application can determine the attribute information of a target object and the target symptom information related to the target object's condition. Then, based on the attribute information and target symptom information, it determines the target level characterizing the severity of the condition and the target guidance information for alleviating the target object's condition. Furthermore, based on the target language processing model, the target level, and the target guidance information, it determines the target feedback information for the target object. In this way, the severity of the condition can be directly determined based on the attribute information and symptom information, and then the rescue vehicles can be rationally allocated according to the severity of the condition, instead of relying entirely on the operator's professional knowledge for vehicle dispatch as in related technologies. Moreover, by combining the target language processing model, the target level, and the guidance information for alleviating the patient's condition to determine the rescue measures for the patient, it not only improves the accuracy of emergency guidance but also ensures that the patient receives timely treatment, solving the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy of emergency guidance in related technologies during emergency dispatch. Attached Figure Description
[0049] Figure 1 A flowchart illustrating an information determination method provided in an embodiment of this application;
[0050] Figure 2 This is a system schematic diagram corresponding to an information determination method provided in an embodiment of this application;
[0051] Figure 3 A flowchart illustrating another information determination method provided in an embodiment of this application;
[0052] Figure 4 This is a text conversion diagram in an information determination method provided in an embodiment of this application;
[0053] Figure 5 This is a schematic diagram of module allocation in an information determination method provided in an embodiment of this application;
[0054] Figure 6 A schematic diagram of an information extraction module in an information determination method provided in an embodiment of this application;
[0055] Figure 7 A schematic diagram of a disease triage module in an information determination method provided in an embodiment of this application;
[0056] Figure 8 A flowchart illustrating a deep learning model in an information determination method provided in this application embodiment;
[0057] Figure 9 A schematic diagram of a target adjustment module in an information determination method provided in an embodiment of this application;
[0058] Figure 10 A schematic diagram of a target classification module in an information determination method provided in an embodiment of this application;
[0059] Figure 11 This is a schematic diagram of the structure of an information determination device provided in an embodiment of this application;
[0060] Figure 12 This is a schematic diagram of the structure of an information determination device provided in an embodiment of this application. Detailed Implementation
[0061] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.
[0062] It should be understood that the phrases "embodiments of this application" or "foreign embodiments" throughout the specification mean that a specific feature, structure, or characteristic related to an embodiment is included in at least one embodiment of this application. Therefore, "embodiments of this application" or "in the foreign embodiments" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.
[0063] Unless otherwise specified, any step in the embodiments of this application performed by the electronic device may be executed by the processor of the electronic device. It is also worth noting that the embodiments of this application do not limit the order in which the electronic device performs the following steps. Furthermore, the methods used to process data in different embodiments may be the same or different methods. It should also be noted that any step in the embodiments of this application can be executed independently by the electronic device; that is, when the electronic device performs any step in the following embodiments, it may not depend on the execution of other steps.
[0064] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.
[0065] This application provides a method that can be applied to an information determining device, as described below. Figure 1 As shown, the method may include the following steps:
[0066] Step 101: Determine the attribute information and target symptom information of the target object.
[0067] Among them, target symptom information is information related to the target subject's condition.
[0068] In this embodiment of the application, the target object may refer to a patient who is initiating an emergency request; the attribute information may include the patient's name, gender, and age; the target symptom information may refer to the patient's current symptoms, such as abdominal pain, headache, and difficulty breathing.
[0069] In this embodiment of the application, if the application is applied to an emergency rescue scenario, it can be achieved through... Figure 2 The speech transcription unit in the system converts the caller's and operator's speech data into text data, which is then processed through... Figure 2 The process control module controls the information extraction unit to process the text data, obtaining the patient's attribute and symptom information, and thus generating the patient's electronic medical record. It should be noted that the person seeking help can be the patient themselves, or their family members, friends, etc.
[0070] Step 102: Based on attribute information and target symptom information, determine the target level of the condition and the target guidance information for the target object.
[0071] Among them, the target level represents the severity of the condition; the target guidance information is used to alleviate the condition of the target individual.
[0072] In the embodiments of this application, it can be achieved through Figure 2 The disease triage module within the task execution unit uses a deep learning model to process attribute information and target symptom information to obtain the disease level. Then, it can be used... Figure 2 The rescue guidance module uses a classification model to process attribute information and target symptom information to obtain the category of the illness, and then determines the guidance information (i.e., target guidance information) for the target object based on the category of the illness.
[0073] It should be noted that, preferably, the condition is usually divided into 4 levels according to the severity. The higher the level, the more serious the condition. That is, level 1 represents a mild condition, while level 4 represents a very serious condition. The target guidance information may include cardiopulmonary resuscitation, the Heimlich maneuver, etc.
[0074] Step 103: Based on the target language processing model, target level, and target guidance information, determine the target feedback information for the target object.
[0075] In this application embodiment, the target language processing model may refer to a large language model (LLM). In one possible implementation, the large language model may refer to the Nine Heavens model.
[0076] In the embodiments of this application, it can be achieved through Figure 2The process control module controls the scheduling module, which uses a large language model to summarize and process the target level and target guidance information to obtain initial feedback information. Then, the process control module can convert the format of the feedback information to obtain the target feedback information.
[0077] The information determination method provided in the embodiments of this application can directly determine the severity of the illness based on attribute information and symptom information, and then rationally allocate rescue vehicles according to the severity of the illness, instead of relying entirely on the operator's professional knowledge for vehicle dispatch as in related technologies. Furthermore, by combining the target language processing model, target level, and guidance information for alleviating the patient's condition to determine the rescue measures for the patient, it not only improves the accuracy of emergency guidance but also ensures that the patient can receive timely treatment. This solves the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy of emergency guidance in the emergency dispatch process of related technologies.
[0078] Based on the foregoing embodiments, embodiments of this application provide an information determination method, referring to... Figure 3 As shown, the method may include the following steps:
[0079] Step 201: The information determination device obtains the dialogue information corresponding to the target object.
[0080] In this application embodiment, the target object may refer to a patient who is initiating an emergency request, such as... Figure 4 As shown, the audio data of the person seeking help can be transcribed into a speech text using a speech transcription unit. Then, the speech text of the person seeking help and the operator's feedback text can be concatenated to obtain dialogue information relevant to the target. It should be noted that the person seeking help can be the patient themselves, or a family member, friend, or other person currently beside the patient.
[0081] Step 202: The information determination device uses a target language processing model to process the dialogue information to obtain the initial attribute information and initial symptom information of the target object.
[0082] In one feasible approach, before processing the dialogue information, such as Figure 5 As shown, the flow control module determines which module to invoke from among the dispatcher module, information extraction module, patient triage module, and rescue guidance module to execute different tasks based on the current dialogue information and the historical dialogue information of emergency patients over a period of time. Specifically, the flow control module can select modules by sending instructions to the large language model, as follows:
[0083] "As a process control module, please select the module to be invoked based on the following dialog:"
[0084] {Historical dialogue information + Current dialogue information};
[0085] The selectable modules are:
[0086] 1. 'Dispatcher Module': Simulates a human operator's conversation with a caller and records the entire dialogue.
[0087] 2. 'Information Extraction Module': Extracts the patient's initial attribute information and initial symptom information from the dialogue information;
[0088] 3. 'Disease Triage Module': Classifies patients' conditions into different levels;
[0089] 4. 'Assistance Guidance Module': Outputs assistance guidance measures based on the patient's gender, age, and symptom information;
[0090] Please provide the input sequence number:
[0091] It should be noted that the instruction can refer to the module number or a custom identifier that is set in advance by the user. It only needs to reflect the information of the module to be called.
[0092] In the embodiments of this application, such as Figure 6 As shown, the flow control module can call the information extraction module to extract keywords from the current dialogue information using a large language model, obtaining the target object's original attribute information and original symptom information. Then, the large language model will verify the extracted original attribute information and original symptom information to obtain initial attribute information and initial symptom information. Specifically, it can determine whether the specific age value is within a reasonable age range (e.g., if the obtained age value is 200, it can be determined that the value is abnormal), and whether contact information appears in the dialogue information.
[0093] It should be noted that the original attribute information may include the patient's name, age, gender, contact information, and dispatch address; the original symptom information may include the patient's current symptom information.
[0094] Step 203: The information determination device verifies the initial attribute information and initial symptom information to obtain attribute information and target symptom information.
[0095] In the embodiments of this application, such as Figure 6 As shown, the initial attribute information and initial symptom information can be validated a second time through the flow control module, and the format of the initial attribute information and initial symptom information can be converted. After that, the converted information can be validated again manually to obtain the attribute information and target symptom information of the target object. In one feasible approach, the initial attribute information and initial symptom information can be converted to JSON format.
[0096] In this embodiment, an electronic medical record can be generated based on the obtained attribute information and target symptom information and stored in a database. It should be noted that the electronic medical record will be displayed on a webpage, and human operators can modify it according to the patient's actual situation.
[0097] Step 204: The information determination device determines the initial level of the condition based on attribute information and target symptom information.
[0098] In this embodiment of the application, the initial level of the disease can be determined based on attribute information, target symptom information, and the target triage table stored in the target database.
[0099] It should be noted that the target triage table can be set according to actual needs. Specifically, patients can be divided into three groups: adults, children, and pregnant women. Then, based on the "Expert Consensus on Emergency Triage", "Standards and Interpretation of Obstetric Emergency Triage", and "Standards for Emergency Triage (Adult Part)", the target triage table shown in Table 1 below can be set.
[0100]
[0101] Table 1
[0102] In Table 1, for the same population, different symptoms correspond to different disease levels; however, for different populations, even if they are at the same level, the symptom information is different.
[0103] In the embodiments of this application, step 204 can be implemented by steps 204a to 204c.
[0104] Step 204a: The information determination device determines the first category of the target object based on the attribute information and the target symptom information.
[0105] In this embodiment of the application, the first category can be an adult, a child, or a pregnant woman. Specifically, as shown... Figure 7 As shown, the patient triage module can input gender, age, and target symptom information from the attribute information into the population classification model to obtain the category to which the patient (i.e., the target object) belongs. It should be noted that the population classification model can refer to the Bidirectional Encoder Representations from Transformers (BERT) model.
[0106] Step 204b: If, based on the mapping relationship between the first category and the target, it is determined that there is matching symptom information in the target database that matches the target symptom information, the information determining device determines the level corresponding to the matching symptom information as the initial level.
[0107] The target mapping relationship is the mapping relationship between disease category and symptom information.
[0108] In this embodiment of the application, the target mapping relationship can refer to the correspondence between different population groups and symptom information in the target triage table. Specifically, such as... Figure 7 As shown, based on the target correspondence, symptom information matching the first category can be determined from the target triage table stored in the target database as the symptom information to be screened. For example, if the first category is determined to be adults, then all symptom information corresponding to the column for adults in the target triage table are the symptom information to be screened.
[0109] Next, the symptom information to be screened can be matched with the target symptom information. If there is a symptom information in the symptom information to be screened that matches the target symptom information (i.e., the matching symptom information), then the level corresponding to the matching symptom information in the target triage table can be determined as the initial level.
[0110] For example, if the target subject is an adult and the target symptom information is asthma and persistent vomiting, it can be determined that there is symptom information matching the target symptom information in the target triage table stored in the target database. In this case, the level corresponding to the symptom information is determined, i.e., level 3 is the initial level. Correspondingly, if the target symptom information is severe dyspnea and asthma, it can be determined that the level corresponding to severe dyspnea is level 2 and the level corresponding to asthma is level 3. In this case, the level with the highest severity of the condition is determined from multiple levels as the initial level, i.e., level 2 is determined as the initial level of the target subject's condition.
[0111] Step 204c: If it is determined that there is no matching symptom information in the target database, the information determination device uses the target grading model to process the attribute information and target symptom information to obtain the initial level.
[0112] In this embodiment, the target hierarchical model can refer to a deep learning model. Specifically, the target hierarchical model may include BERT layers, feedforward network (FN) layers, and normalization layers (hereinafter referred to as: Softmax layers). The feedforward network model can be constructed from multiple fully connected layers.
[0113] In this embodiment, semantic information can be extracted using target symptom information from the BERT layer. Specifically, the text length of the symptom information can be set to seq_len, and the text size to embedding_size. The size of the extracted feature vector of the target symptom information is then seq_len * embedding_size. Gender can be represented by a one-dimensional vector of size 100, with males represented by vectors of all 1s, females by vectors of all 2s, and unknown genders by vectors of all 0s. Age can also be represented by a one-dimensional vector of size 100. Specifically, age can be represented as an integer, and then the index of the vector corresponding to the specific age value can be set to 1. For example, if the age value is 50, the bit at index 50 in the vector is represented by 1, and the rest by 0. Furthermore, if the age value is greater than or equal to 100, the bit at index 100 in the vector is represented by 1, and the first 99 bits are represented by 0.
[0114] Then, the age vector and gender vector can be concatenated according to formula (1) to extract the semantic features of the target symptom information, resulting in a 2*embedding_size feature vector. This 2*embedding_size feature vector is then concatenated with a feature vector of size seq_len*embedding_size to obtain a target feature vector of size (seq_len+2)*embedding_size. It should be noted that formula (1) can be expressed as follows:
[0115]
[0116] In this context, [1,1,1...1...1,1,1,1] represents gender, where all digits are 1, indicating the patient is male; [0,0,1...0...0,0,0,0] represents age, where the third digit is 1, indicating the patient is 3 years old.
[0117] Furthermore, it can be like Figure 8 As shown, the obtained target feature vector is sequentially input into the feedforward network model and the Softmax layer to obtain the probability of the patient's (i.e. the target object) disease level for each level. Then, the level with the highest probability is determined as the initial level.
[0118] For example, if the probability of a patient's condition being grade 1 is 20%, grade 2 is 40%, grade 3 is 35%, and grade 4 is 5%, then the initial grade can be determined to be grade 4.
[0119] Step 205: The information determination device acquires the target quantity of objects to be processed at the initial level and the target parameters of the rescue vehicle.
[0120] Among them, the target parameters characterize the usage of rescue vehicles.
[0121] In this embodiment of the application, the target number of objects to be processed may refer to the number of patients at the initial level among those waiting for assistance; the target parameters may include the total number of rescue vehicles and the number already used.
[0122] Step 206: The information determination device uses a target adjustment model to process the target symptom information, target quantity, and target parameters to obtain the target level.
[0123] In this embodiment of the application, the target adjustment model can be trained in the following way:
[0124] A1. The information determination device obtains historical attribute information and historical symptom information of historical objects.
[0125] Among them, historical symptom information is information related to the patient's medical condition in the past.
[0126] In this embodiment of the application, the historical object may refer to a patient who initiated an emergency request within a certain period of time in the past; historical attribute information may include information such as the patient's age and gender.
[0127] A2. The information determination device determines the historical level of a historical subject's condition based on historical attribute information and historical symptom information.
[0128] In this embodiment, the historical level can refer to the severity of a patient's condition who initiated an emergency request within a past period. Specifically, step 204 can be used to determine the historical level of the condition based on attribute information, historical symptom information, and the target triage table.
[0129] It should be noted that although both historical level and initial level refer to the severity of a patient's condition, the initial level refers to the severity of the patient currently making an emergency request, while the historical level refers to the severity of the patient's condition in the past. The two are not related.
[0130] For example, if two patients received assistance in the past hour, their condition levels can be determined to be Level 1 and Level 3 based on their attribute and symptom information. Level 1 and Level 3 are the historical levels of their conditions.
[0131] A3. The information determination equipment trains the initial adjustment model based on historical quantity, historical parameters of rescue vehicles, and historical symptom information to obtain the target adjustment model.
[0132] Among them, the historical quantity refers to the number of historical objects at the historical level; the historical parameters characterize the historical usage of rescue vehicles.
[0133] In the embodiments of this application, such as Figure 9 As shown, the initial adjustment model can be used to adjust the initial level of historical objects and consists of a representation layer and a fusion layer; wherein, the representation layer may include a BERT layer, a bidirectional long short-term memory (BiLSTM) layer and a first fully connected layer (FC); the fusion layer includes a second fully connected layer and a third fully connected layer.
[0134] In this embodiment, historical data, historical parameters, and historical symptom information can be input into the initial adjustment model to obtain the prediction level. Then, based on the prediction level and the determined historical level, the loss value of the initial adjustment model is determined, and the parameters of the initial adjustment model are updated based on the loss value. Then, when the accuracy of the prediction level meets the preset threshold, training is stopped, and the initial adjustment model at this moment is determined as the target adjustment model.
[0135] In this embodiment of the application, by training the initial adjustment model based on the symptom information, historical level, historical quantity, and historical parameters of historical subjects, an accurate level adjustment model can be obtained to accurately determine the level of the patient's condition and to reasonably allocate rescue vehicles.
[0136] In the embodiments of this application, such as Figure 9 As shown, target symptom information, target quantity, and target parameters can be input into the target adjustment model. At this point, the BERT layer in the representation layer can extract semantic features from the target symptom information to obtain a semantic feature sequence. Then, a BiLSTM layer extracts features from the semantic feature sequence to obtain the feature information v of the target symptom information. (t) Simultaneously, the two first fully connected layers in the representation layer can extract feature information v regarding the number of targets and target parameters. (s) Then, the second fully connected layer in the fusion layer performs dimension alignment on the two obtained feature information, resulting in two processed feature information. and Furthermore, the two processed feature information are concatenated by the concatenation module, and the concatenated feature is input into the third fully connected layer to obtain the probability of the target object's condition at each level. Then, the level with the highest probability is determined as the final level of the target object's condition, i.e., the target level.
[0137] It should be noted that after obtaining the target level, the data format corresponding to the target level needs to be validated through the process control module.
[0138] In this embodiment, the patient's condition can be classified based on the patient's symptoms, and the patient's classification can be adjusted based on the resources available at the hospital, such as the number of ambulances, the number of ambulances already dispatched, and the number of patients at the current level, to obtain the optimal classification result. This assists the dispatch center in the intelligent and rational allocation of resources, so as to prioritize dispatching ambulances to patients with urgent conditions in situations where emergency resources are scarce (during large-scale disasters), thereby improving emergency response efficiency.
[0139] Step 207: The information determination device uses a target classification model to process the attribute information and target symptom information to obtain the target category of the disease.
[0140] In this embodiment of the application, the target classification model can be used to classify the patient's condition, and the target classification model can be trained in the following way:
[0141] B1. Information determination device obtains historical dialogue information corresponding to historical objects.
[0142] In this embodiment of the application, historical dialogue information between patients and operators over a past period can be obtained from the database.
[0143] B2. The information determination device uses a target language processing model to process historical dialogue information to obtain historical attribute information and historical symptom information.
[0144] In this embodiment of the application, keywords can be extracted from historical dialogue information to obtain the attribute information and symptom information of historical objects. Then, the large language model will perform verification processing on the extracted attribute information and symptom information to obtain historical attribute information and historical symptom information.
[0145] B3. The information determination device trains the initial classification model based on historical attribute information and historical symptom information to obtain the target classification model.
[0146] In this embodiment, the initial classification model may include a BERT layer, a fully connected layer, and a Softmax layer. Specifically, historical attribute information and historical symptom information can be input into the initial classification model to predict the category of the condition. Then, based on the historical category and the predicted category, the loss value of the initial classification model is determined, and the parameters of the initial classification model are updated based on the loss value. Finally, training stops when the accuracy of the predicted category meets a threshold, so as to obtain the target classification model.
[0147] In the embodiments of this application, such as Figure 10As shown, attribute information and target symptom information can be input into the target classification model to obtain the probability that the target object's disease belongs to each disease in the International Classification of Diseases (ICD) system. Then, the category of the disease with the highest probability is determined as the target category of the target object's condition. It should be noted that there are multiple versions of the International Classification of Diseases system; preferably, ICD-10 can be used.
[0148] It should be noted that step 208 or step 209 can be executed after step 207.
[0149] Step 208: If a second category matching the target category exists in the disease categories stored in the target database, the information determining device determines the guidance information corresponding to the second category as the target guidance information.
[0150] In this embodiment, the target database may refer to an emergency medical knowledge base. Specifically, if a second category matching the target category exists in the target database, the emergency medical guidance measures (i.e., guidance information) corresponding to that second category can be identified as the target guidance information. For example, if the identified target category is the respiratory system, it can be determined whether information on respiratory system diseases exists in the target database. If so, the emergency medical guidance measures for the respiratory system can be identified as the target guidance information.
[0151] In one feasible approach, if multiple pieces of guidance information are obtained, they can be aggregated and processed using a large language model to obtain the target guidance information.
[0152] Step 209: If there is no second category in the disease category, the information determination device uses the target language processing model to process the target category, attribute information and symptom information to obtain target guidance information.
[0153] In this embodiment, if no second category matching the target category exists in the target database, a large language model can be used to process the target category, attribute information, and target symptom information to generate guidance information (target guidance information) for the target object. It should be noted that the generated guidance information can be manually modified according to the patient's actual situation, and regardless of how the target guidance information is obtained, it needs to be manually verified a second time to ensure its accuracy. It should also be noted that after obtaining the target guidance information, the format of the target guidance information needs to be verified through a process control module.
[0154] In this embodiment, corresponding guidance measures can be retrieved from the emergency knowledge base based on the patient's condition level. If no results are found, emergency guidance measures for the patient can be generated directly based on the capabilities of the large language model. This ensures that the emergency guidance measures are perfectly matched with the patient's condition, thereby improving the efficiency of patient rescue.
[0155] Step 210: The information determination device uses the target language processing model to summarize and process the target level and target guidance information to obtain initial feedback information for the target object.
[0156] In this embodiment of the application, the process control module can call the scheduler module to input the target level and target guidance information into the large language model, so that the large language model can summarize and process the input information to obtain initial feedback information.
[0157] It should be noted that the initial feedback information is in text format.
[0158] Step 211: The information determination device converts the format of the initial feedback information to obtain the target feedback information.
[0159] In this embodiment of the application, after obtaining the initial feedback information, the initial feedback information can be verified by the process control module. Then, the initial feedback information in text format will be converted into target feedback information in audio format and broadcast to the person seeking help.
[0160] In this embodiment of the application, after the scheduling task is completed, the obtained electronic medical records, disease grading results and emergency guidance measures can be verified, and the model used can be updated based on the verified information to improve the accuracy of the model.
[0161] In this embodiment, the dispatcher module can generate target feedback information and provide voice feedback to the person seeking help. This can provide accurate first aid measures to the person seeking help or the patient in a concise manner, while also calming the person seeking help or the patient, so as to ensure that the person seeking help or the patient can receive timely assistance.
[0162] In other embodiments of this application, step 212 may also be performed after step 211.
[0163] Step 212: The information determination device updates the target database based on the dialogue information and target feedback information corresponding to the target object.
[0164] In this embodiment of the application, after the target feedback information is determined, the dialogue information and the target feedback information can be stored in the database in a one-to-one correspondence.
[0165] The information determination method provided in the embodiments of this application can directly determine the severity of the illness based on attribute information and symptom information, and then rationally allocate rescue vehicles according to the severity of the illness, instead of relying entirely on the operator's professional knowledge for vehicle dispatch as in related technologies. Furthermore, by combining the target language processing model, target level, and guidance information for alleviating the patient's condition to determine the rescue measures for the patient, it not only improves the accuracy of emergency guidance but also ensures that the patient can receive timely treatment. This solves the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy of emergency guidance in the emergency dispatch process of related technologies.
[0166] Based on the foregoing embodiments, this application provides an information determining device, which can be applied to... Figure 1 and 3 In the information determination method provided in the corresponding embodiment, refer to Figure 11 As shown, the information determining device 3 may include: a first determining unit 31, a second determining unit 32, and a third determining unit 33, wherein:
[0167] The first determining unit 31 is used to determine the attribute information and target symptom information of the target object; wherein, the target symptom information is information related to the condition of the target object;
[0168] The second determining unit 32 is used to determine the target level of the illness and the target guidance information for the target object based on attribute information and target symptom information; wherein, the target level represents the severity of the illness; and the target guidance information is used to alleviate the illness of the target object.
[0169] The third determining unit 33 is used to determine the target feedback information for the target object based on the target language processing model, target level and target guidance information.
[0170] In other embodiments of this application, the first determining unit 31 is further configured to perform the following steps:
[0171] Retrieve the dialogue information corresponding to the target object;
[0172] The target language processing model is used to process the dialogue information to obtain the initial attribute information and initial symptom information of the target object;
[0173] The initial attribute information and initial symptom information are validated to obtain the attribute information and target symptom information.
[0174] In other embodiments of this application, the second determining unit 32 is further configured to perform the following steps:
[0175] Based on attribute information and target symptom information, determine the initial level of the disease;
[0176] Obtain the target number and target parameters of the objects to be processed at the initial level; where the target parameters represent the usage status of the rescue vehicles;
[0177] A target adjustment model is used to process target symptom information, target quantity, and target parameters to obtain target levels.
[0178] In other embodiments of this application, the second determining unit 32 is further configured to perform the following steps:
[0179] Based on attribute information and target symptom information, the first category of the target object is determined;
[0180] If, based on the first category and the target mapping relationship, it is determined that there is matching symptom information in the target database that matches the target symptom information, the level corresponding to the matching symptom information is determined as the initial level; where the target mapping relationship is the mapping relationship between the disease category and the symptom information;
[0181] If it is determined that no matching symptom information exists in the target database, the attribute information and target symptom information are processed using a target grading model to obtain an initial level.
[0182] In other embodiments of this application, the second determining unit 32 is further configured to perform the following steps:
[0183] Obtain historical attribute information and historical symptom information of historical objects; among which, historical symptom information is information related to the illness of historical objects.
[0184] Based on historical attribute information and historical symptom information, determine the historical level of the patient's condition;
[0185] Based on historical quantity, historical parameters, and historical symptom information, the initial adjustment model is trained to obtain the target adjustment model; where historical quantity refers to the number of historical objects at the historical level; historical parameters characterize the historical usage of rescue vehicles.
[0186] In other embodiments of this application, the second determining unit 32 is further configured to perform the following steps:
[0187] A target classification model is used to process attribute information and target symptom information to obtain the target category of the disease.
[0188] If a second category matching the target category exists in the disease categories stored in the target database, the guidance information corresponding to the second category is determined as the target guidance information;
[0189] If there is no second category in the disease category, the target language processing model is used to process the target category, attribute information and symptom information to obtain target guidance information.
[0190] In other embodiments of this application, the second determining unit 32 is further configured to perform the following steps:
[0191] Retrieve historical dialogue information corresponding to historical objects;
[0192] The target language processing model is used to process historical dialogue information to obtain historical attribute information and historical symptom information;
[0193] Based on historical attribute information and historical symptom information, the initial classification model is trained to obtain the target classification model.
[0194] In other embodiments of this application, the third determining unit 33 is further configured to perform the following steps:
[0195] The target language processing model is used to summarize and process the target level and target guidance information to obtain initial feedback information for the target object;
[0196] The initial feedback information is converted into the target feedback information.
[0197] In other embodiments of this application, the third determining unit 33 is further configured to perform the following steps:
[0198] Update the target database based on the dialogue information and target feedback information corresponding to the target object.
[0199] It should be noted that a detailed explanation of the steps performed by each unit can be found in [reference needed]. Figure 1 and 3 The information determination method provided in the corresponding embodiments will not be described again here.
[0200] The information determination device provided in the embodiments of this application can directly determine the severity of the illness based on attribute information and symptom information, and then rationally allocate rescue vehicles according to the severity of the illness, instead of relying entirely on the operator's professional knowledge for vehicle dispatch as in related technologies. Furthermore, by combining the target language processing model, target level, and guidance information for alleviating the patient's condition to determine the rescue measures for the patient, it not only improves the accuracy of emergency guidance but also ensures that the patient can receive timely treatment. This solves the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy of emergency guidance in the emergency dispatch process of related technologies.
[0201] Based on the foregoing embodiments, embodiments of this application provide an information determining device, which can be applied to... Figure 1 and 3 In the information determination method provided in the corresponding embodiment, refer to Figure 12As shown, the information determining device 4 may include: a processor 41, a memory 42, and a communication bus 43, wherein:
[0202] Communication bus 43 is used to realize the communication connection between processor 41 and memory 42;
[0203] The processor 41 is used to execute the information determination program in the memory 42 to perform the following steps:
[0204] Determine the target object's attribute information and target symptom information; among which, the target symptom information is information related to the target object's condition.
[0205] Based on attribute information and target symptom information, the target level of the illness and the target guidance information for the target subject are determined; whereby the target level represents the severity of the illness; and the target guidance information is used to alleviate the illness of the target subject.
[0206] Based on the target language processing model, target level, and target guidance information, target feedback information is determined for the target object.
[0207] In other embodiments of this application, the processor 41 is used to execute the attribute information and target symptom information of the target object stored in the memory 42 to implement the following steps:
[0208] Retrieve the dialogue information corresponding to the target object;
[0209] The target language processing model is used to process the dialogue information to obtain the initial attribute information and initial symptom information of the target object;
[0210] The initial attribute information and initial symptom information are validated to obtain the attribute information and target symptom information.
[0211] In other embodiments of this application, the processor 41 is used to execute the information determination program in the memory 42 to determine the target level of the condition based on attribute information and target symptom information, in order to implement the following steps:
[0212] Based on attribute information and target symptom information, determine the initial level of the disease;
[0213] Obtain the target number and target parameters of the objects to be processed at the initial level; where the target parameters represent the usage status of the rescue vehicles;
[0214] A target adjustment model is used to process target symptom information, target quantity, and target parameters to obtain target levels.
[0215] In other embodiments of this application, the processor 41 is used to execute the information determination program in the memory 42 to determine the initial level of the condition based on attribute information and target symptom information, in order to implement the following steps:
[0216] Based on attribute information and target symptom information, the first category of the target object is determined;
[0217] If, based on the first category and the target mapping relationship, it is determined that there is matching symptom information in the target database that matches the target symptom information, the level corresponding to the matching symptom information is determined as the initial level; where the target mapping relationship is the mapping relationship between the disease category and the symptom information;
[0218] If it is determined that no matching symptom information exists in the target database, the attribute information and target symptom information are processed using a target grading model to obtain an initial level.
[0219] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps:
[0220] Obtain historical attribute information and historical symptom information of historical objects; among which, historical symptom information is information related to the illness of historical objects.
[0221] Based on historical attribute information and historical symptom information, determine the historical level of the patient's condition;
[0222] Based on historical quantity, historical parameters, and historical symptom information, the initial adjustment model is trained to obtain the target adjustment model; where historical quantity refers to the number of historical objects at the historical level; historical parameters characterize the historical usage of rescue vehicles.
[0223] In other embodiments of this application, processor 41 is used to execute the information determination program in memory 42 to determine target guidance information for the target object based on attribute information and target symptom information, in order to implement the following steps:
[0224] A target classification model is used to process attribute information and target symptom information to obtain the target category of the disease.
[0225] If a second category matching the target category exists in the disease categories stored in the target database, the guidance information corresponding to the second category is determined as the target guidance information;
[0226] If there is no second category in the disease category, the target language processing model is used to process the target category, attribute information and symptom information to obtain target guidance information.
[0227] In other embodiments of this application, the processor 41 is used to execute an information determination program in the memory 42 to perform the following steps:
[0228] Retrieve historical dialogue information corresponding to historical objects;
[0229] The target language processing model is used to process historical dialogue information to obtain historical attribute information and historical symptom information;
[0230] Based on historical attribute information and historical symptom information, the initial classification model is trained to obtain the target classification model.
[0231] In other embodiments of this application, the processor 41 is further configured to execute the transfer information determination program in the memory 42 based on the target language processing model, target level, and target guidance information to determine target feedback information for the target object, in order to implement the following steps:
[0232] The target language processing model is used to summarize and process the target level and target guidance information to obtain initial feedback information for the target object;
[0233] The initial feedback information is converted into the target feedback information.
[0234] In other embodiments of this application, the processor 41 is further configured to execute a transmission information determination program in the memory 42 to perform the following steps:
[0235] Update the target database based on the dialogue information and target feedback information corresponding to the target object.
[0236] It should be noted that a detailed description of the steps performed by the processor can be found in [reference needed]. Figure 1 and 3 The information determination method provided in the corresponding embodiments will not be described again here.
[0237] The information determination device provided in the embodiments of this application can directly determine the severity of a patient's condition based on attribute information and symptom information, and then rationally allocate rescue vehicles according to the severity of the condition, instead of relying entirely on the operator's professional knowledge for vehicle dispatch as in related technologies. Furthermore, by combining the target language processing model, target level, and guidance information for alleviating the patient's condition to determine the rescue measures for the patient, it not only improves the accuracy of emergency guidance but also ensures that the patient can receive timely treatment. This solves the problems of poor accuracy and low efficiency in vehicle allocation and low accuracy of emergency guidance in related technologies during emergency dispatch.
[0238] Based on the foregoing embodiments, embodiments of this application provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to achieve... Figure 1 and 3 The corresponding embodiments provide the steps of the information determination method.
[0239] Based on the foregoing embodiments, embodiments of this application provide a computer program product, which includes a computer program that, when executed by a processor, implements... Figure 1 and 3 The corresponding embodiments provide the steps of the information determination method.
[0240] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.
[0241] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0242] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0243] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0244] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A method for determining information, characterized in that, The method includes: Determine the attribute information and target symptom information of the target object; wherein, the target symptom information is information related to the condition of the target object; Based on the attribute information and the target symptom information, a target level for the condition and target guidance information for the target individual are determined; wherein, the target level characterizes the severity of the condition; and the target guidance information is used to alleviate the condition of the target individual. Based on the target language processing model, the target level, and the target guidance information, target feedback information is determined for the target object.
2. The method according to claim 1, characterized in that, The determination of the target object's attribute information and target symptom information includes: Obtain the dialogue information corresponding to the target object; The target language processing model is used to process the dialogue information to obtain the initial attribute information and initial symptom information of the target object; The initial attribute information and the initial symptom information are validated to obtain the attribute information and the target symptom information.
3. The method according to claim 1, characterized in that, Based on the attribute information and the target symptom information, the target level of the condition is determined, including: Based on the attribute information and the target symptom information, the initial level of the condition is determined; Obtain the target number and target parameters of the objects to be processed at the initial level; wherein, the target parameters characterize the usage status of the rescue vehicle; The target symptom information, the number of targets, and the target parameters are processed using a target adjustment model to obtain the target level.
4. The method according to claim 3, characterized in that, Determining the initial level of the condition based on the attribute information and the target symptom information includes: Based on the attribute information and the target symptom information, the first category of the target object is determined; If, based on the first category and the target mapping relationship, it is determined that there is matching symptom information in the target database that matches the target symptom information, the level corresponding to the matching symptom information is determined as the initial level; wherein, the target mapping relationship is the mapping relationship between disease category and symptom information; If it is determined that the matching symptom information does not exist in the target database, the attribute information and the target symptom information are processed using a target grading model to obtain the initial level.
5. The method according to claim 4, characterized in that, The method further includes: Obtain historical attribute information and historical symptom information of the historical object; wherein, the historical symptom information is information related to the illness of the historical object; Based on the historical attribute information and the historical symptom information, the historical level of the patient's condition is determined; Based on historical data, historical parameters, and the historical symptom information, the initial adjustment model is trained to obtain the target adjustment model; wherein, the historical data refers to the number of historical objects at the historical level; and the historical parameters characterize the historical usage of the rescue vehicle.
6. The method according to claim 1, characterized in that, Based on the attribute information and the target symptom information, target guidance information is determined for the target object, including: The attribute information and the target symptom information are processed using a target classification model to obtain the target category of the disease. If a second category matching the target category exists in the disease categories stored in the target database, the guidance information corresponding to the second category is determined as the target guidance information; If the second category does not exist in the disease category, the target language processing model is used to process the target category, the attribute information, and the symptom information to obtain the target guidance information.
7. The method according to claim 6, characterized in that, The method further includes: Retrieve historical dialogue information corresponding to historical objects; The target language processing model is used to process the historical dialogue information to obtain the historical attribute information and the historical symptom information; Based on the historical attribute information and the historical symptom information, the initial classification model is trained to obtain the target classification model.
8. The method according to claim 1, characterized in that, The determination of target feedback information for the target object based on the target language processing model, the target level, and the target guidance information includes: The target language processing model is used to summarize and process the target level and the target guidance information to obtain initial feedback information for the target object; The initial feedback information is converted to obtain the target feedback information; Accordingly, the method further includes: The target database is updated based on the dialogue information corresponding to the target object and the target feedback information.
9. An information determining device, characterized in that, The device includes: The first determining unit is used to determine the attribute information and target symptom information of the target object; wherein the target symptom information is information related to the condition of the target object; The second determining unit is configured to determine the target level of the illness and target guidance information for the target object based on the attribute information and the target symptom information; wherein the target level characterizes the severity of the illness; and the target guidance information is used to alleviate the illness of the target object. The third determining unit is used to determine target feedback information for the target object based on the target language processing model, the target level, and the target guidance information.
10. An information determining device, characterized in that, The device includes: a processor, a memory, and a communication bus; The communication bus is used to realize the communication connection between the processor and the memory; The processor is used to execute an information determination program in the memory to implement the steps of the information determination method as described in any one of claims 1 to 8.
11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the information determination method as described in any one of claims 1 to 8.
12. A computer program product, the computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the information determination method according to any one of claims 1 to 8.