Diagnosis and treatment data analysis method and device for chest pain patient

Abstract

The invention relates to the technical field of medical data processing, and provides a diagnosis and treatment data analysis method and device for a chest pain patient, and the method comprises the steps: obtaining target electrocardiogram data, patient medical history information, past medical history and medication history, medical equipment position information, and a target symptom signal collected by a symptom monitoring device; utilizing a chest pain symptom feature extraction model, a chest pain medical history feature extraction model and a symptom signal feature extraction model in the chest pain diagnosis and treatment analysis model to respectively extract chest pain symptom features such as electrocardiogram waveform features and pain part features, chest pain medical history features and symptom signal features; and finally, combining the position information and the characteristics to generate chest pain diagnosis and treatment information, including chest pain types, symptom description and medical history records. According to the method, multi-dimensional data fusion analysis is realized, and accurate decision support is provided for chest pain diagnosis and treatment. The diagnosis efficiency and accuracy can be improved, and dynamic update of the diagnosis and treatment information base and optimization of first-aid resource configuration are supported.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data processing, and more specifically, to a method and device for analyzing diagnosis and treatment data of chest pain patients. Background Art

[0002] In the modern medical field, with the continuous increase in the incidence of cardiovascular diseases, the diagnosis and treatment of chest pain patients have become an important topic in clinical medicine. Chest pain is a common and complex symptom that may be caused by various reasons, including but not limited to serious diseases such as coronary heart disease, myocardial infarction, angina pectoris, and pulmonary embolism. Accurately and quickly diagnosing and treating chest pain patients is of crucial significance for reducing the mortality rate of patients and improving the prognosis.

[0003] Traditional methods for diagnosing chest pain mainly rely on doctors' subjective judgment of patients' symptoms, electrocardiogram (ECG) examinations, and medical history inquiries. However, this method has some limitations. First, the experience and knowledge level of doctors have a greater impact on the diagnosis results, which may lead to insufficient accuracy and consistency of the diagnosis. Second, although the ECG examination is an important diagnostic tool, its sensitivity and specificity for certain chest pain etiologies are limited, especially in cases of atypical symptoms or early stages. In addition, the collection and analysis of medical history information often take a lot of time and are easily affected by patients' memory biases or inaccurate expressions. In the actual medical environment, especially in emergency situations, quickly and accurately obtaining and analyzing this information is crucial for timely treating patients.

[0004] In recent years, with the continuous development of information technology and medical technology, methods for using computer-aided diagnosis systems to analyze chest pain patient data have emerged. These systems usually combine multi-source data such as ECG data and medical history information, and use certain algorithms or models to assist doctors in making diagnoses. However, most of these systems have the following problems: on the one hand, the analysis of ECG data is often limited to simple feature extraction and fails to fully utilize the complex information in the ECG; on the other hand, the processing methods of medical history information and symptom signals are relatively single, lacking in-depth mining and integration of multi-dimensional data. In addition, most of these systems do not consider the geographical location information of medical devices, which is of great significance for optimizing the allocation of first aid resources and improving first aid efficiency.

[0005] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the prior art: the prior art cannot fully mine and integrate the complex information in multi-source data, resulting in limited diagnostic accuracy and efficiency; the lack of utilization of the geographical location information of medical devices makes it impossible to effectively optimize the allocation of first aid resources; the processing methods of medical history information and symptom signals are relatively simple and cannot meet the diagnostic requirements in complex clinical scenarios. Summary of the Invention

[0006] The present invention provides a method and device for analyzing diagnosis and treatment data of chest pain patients.

[0007] In a first aspect of the present invention, there is provided a method for analyzing diagnosis and treatment data of chest pain patients, including:

[0008] Obtain target electrocardiogram data, patient medical history information, location information, and target symptom signals, wherein the patient medical history information includes: past medical history records and medication history information, the location information represents the installation location of the medical device that collects the target electrocardiogram data, and the target symptom signals are collected by a symptom monitoring device;

[0009] Extract chest pain symptom features from the target electrocardiogram data through a chest pain symptom feature extraction model to generate chest pain symptom features, wherein the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features;

[0010] Extract chest pain medical history features from the patient medical history information through a chest pain medical history feature extraction model to generate chest pain medical history features;

[0011] Extract symptom signal features from the target symptom signals through a symptom signal feature extraction model to generate symptom signal features, wherein the chest pain symptom feature extraction model, the chest pain medical history feature extraction model, and the symptom signal feature extraction model are included in a chest pain diagnosis and treatment analysis model;

[0012] Determine chest pain diagnosis and treatment information according to the location information, the chest pain symptom features, the chest pain medical history features, and the symptom signal features, wherein the chest pain diagnosis and treatment information includes: location information, chest pain type, chest pain symptom description, and chest pain medical history record.

[0013] Further, the method further includes:

[0014] Determine whether there is a diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in a chest pain diagnosis and treatment information database according to the chest pain type included in the chest pain diagnosis and treatment information;

[0015] In response to determining that there is no diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database, add the chest pain diagnosis and treatment information as a diagnosis and treatment information record to the chest pain diagnosis and treatment information database;

[0016] In response to determining that there is a diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database, update the diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information according to the chest pain diagnosis and treatment information.

[0017] Further, the method further includes:

[0018] Read the updated chest pain diagnosis and treatment information database to obtain a set of diagnosis and treatment information records;

[0019] Determine the current chest pain distribution map according to the set of diagnosis and treatment information records;

[0020] Obtain the historical chest pain distribution map;

[0021] Generate chest pain trend information according to the current chest pain distribution map and the historical chest pain distribution map.

[0022] Further, the method further includes:

[0023] Mark the first aid equipment to be updated in position according to the current chest pain distribution map and the chest pain trend information to generate marked first aid equipment information, and obtain a set of marked first aid equipment information;

[0024] Optimize the set position of the first aid equipment corresponding to each piece of marked first aid equipment information in the set of marked first aid equipment information.

[0025] Further, the chest pain symptom feature extraction model includes: a basic feature generator, a complex feature encoder, a spatial anchor point generator, a feature area calibrator, a first set of local feature extraction networks, a second set of local feature extraction networks, an electrocardiogram waveform feature extraction layer, a pain location feature extraction layer, a pain nature feature extraction layer, an accompanying symptom feature extraction layer, and an onset time feature extraction layer. The network structures of the spatial anchor point generator and the feature area calibrator are the same; and the extraction of chest pain symptom features from the target electrocardiogram data through the chest pain symptom feature extraction model to generate chest pain symptom features includes:

[0026] Extract basic features from the target electrocardiogram data through the basic feature generator to generate a basic electrocardiogram feature map;

[0027] Perform complex feature encoding on the basic electrocardiogram feature map through the complex feature encoder to generate a set of complex electrocardiogram feature maps;

[0028] Generate a set of spatial anchor point feature maps through the spatial anchor point generator and the basic electrocardiogram feature map;

[0029] Generate a set of calibrated feature maps through the feature area calibrator and the target complex electrocardiogram feature map, where the target complex electrocardiogram feature map is the complex electrocardiogram feature map in the set of complex electrocardiogram feature maps with a target ratio of the image size corresponding to the size of the target electrocardiogram data;

[0030] Through the first set of local feature extraction networks, perform local feature extraction on the spatial anchor feature maps in the set of spatial anchor feature maps to generate first local feature maps, obtaining a set of first local feature maps;

[0031] Through the second set of local feature extraction networks, perform local feature extraction on the calibrated feature maps in the set of calibrated feature maps to generate second local feature maps, obtaining a set of second local feature maps;

[0032] Concatenate and fuse the set of first local feature maps, the set of second local feature maps, and the complex electrocardiogram feature maps in the set of complex electrocardiogram feature maps except for the target complex electrocardiogram feature map to obtain a concatenated and fused feature map;

[0033] Input the concatenated and fused feature map into the electrocardiogram waveform feature extraction layer to obtain the electrocardiogram waveform feature;

[0034] Input the concatenated and fused feature map into the pain location feature extraction layer to obtain the pain location feature;

[0035] Input the concatenated and fused feature map into the pain nature feature extraction layer to obtain the pain nature feature;

[0036] Input the concatenated and fused feature map into the accompanying symptom feature extraction layer to obtain the accompanying symptom feature;

[0037] Input the concatenated and fused feature map into the attack time feature extraction layer to obtain the attack time feature.

[0038] Furthermore, the chest pain medical history feature extraction model includes: a first feature transformation layer, a medical history multi-dimensional feature extraction network, a second feature transformation layer, a medical history dynamic feature extraction network, a feature aggregation layer, a first feature mapping layer, and a second feature mapping layer; and through the chest pain medical history feature extraction model, perform chest pain medical history feature extraction on the patient's medical history information to generate chest pain medical history features, including:

[0039] Through the first feature transformation layer, perform feature transformation processing on the patient's medical history information to generate a first transformed medical history feature;

[0040] Through the medical history multi-dimensional feature extraction network, perform multi-dimensional feature extraction on the first transformed medical history feature to obtain multi-dimensional medical history features;

[0041] Through the second feature transformation layer, perform feature transformation processing on the multi-dimensional medical history features to generate second transformed medical history features;

[0042] Through the historical medical record dynamic feature extraction network, perform dynamic feature extraction on the second transformed historical medical record features to obtain dynamic historical medical record features;

[0043] Through the feature aggregation layer, perform feature aggregation processing on the dynamic historical medical record features to generate the aggregated dynamic historical medical record features;

[0044] Through the first feature mapping layer, perform a first feature mapping on the aggregated dynamic historical medical record features to obtain the first mapped dynamic historical medical record features;

[0045] Through the second feature mapping layer, perform a second feature mapping on the first mapped dynamic historical medical record features to obtain the chest pain historical medical record features.

[0046] Further, the medication history information includes: a set of medication component description information, and the medication component information in the medication component information set includes: medication component type and dosage, and the symptom signal feature extraction model includes: a symptom multi-dimensional feature extractor and a feature fusion mapping layer; and the process of performing symptom signal feature extraction on the target symptom signal through the symptom signal feature extraction model to generate symptom signal features includes:

[0047] Through the symptom multi-dimensional feature extractor, perform multi-dimensional feature extraction on each piece of medication component information in the medication component information set to generate symptom component features, and obtain a set of symptom component features;

[0048] Perform feature fusion on the symptom component features in the set of symptom component features to obtain the fused symptom component features;

[0049] Perform range mapping processing on the symptom signal to generate the range-mapped symptom signal;

[0050] Perform feature splicing on the range-mapped symptom signal and the fused symptom component features to obtain the spliced features;

[0051] Input the spliced features into the feature fusion mapping layer to obtain the symptom signal features.

[0052] In the second aspect of the present invention, there is provided a diagnosis and treatment data analysis device for chest pain patients, including:

[0053] An acquisition unit configured to acquire target electrocardiogram data, patient medical history information, location information, and target symptom signals, where the patient medical history information includes: past medical history records and medication history information, the location information represents the installation location of the medical device that acquires the target electrocardiogram data, and the target symptom signals are acquired by a symptom monitoring device;

[0054] A chest pain symptom feature extraction unit, configured to extract chest pain symptom features from the target electrocardiogram data through a chest pain symptom feature extraction model to generate chest pain symptom features, where the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features;

[0055] A chest pain medical history feature extraction unit, configured to extract chest pain medical history features from the patient medical history information through a chest pain medical history feature extraction model to generate chest pain medical history features;

[0056] A symptom signal feature extraction unit, configured to extract symptom signal features from the target symptom signal through a symptom signal feature extraction model to generate symptom signal features, where the chest pain symptom feature extraction model, the chest pain medical history feature extraction model, and the symptom signal feature extraction model are included in a chest pain diagnosis and treatment analysis model;

[0057] A determination unit, configured to determine chest pain diagnosis and treatment information based on the location information, the chest pain symptom features, the chest pain medical history features, and the symptom signal features, where the chest pain diagnosis and treatment information includes: location information, chest pain type, chest pain symptom description, and chest pain medical history record.

[0058] In a third aspect of the present invention, an electronic device is provided, the electronic device includes: at least one processor, a memory, and an input / output unit; wherein, the memory is used to store a computer program, and the processor is used to call the computer program stored in the memory to execute the method according to any one of the first aspect.

[0059] In a fourth aspect of the present invention, a computer-readable storage medium is provided, which includes instructions that, when run on a computer, cause the computer to execute the method according to any one of the first aspect.

[0060] According to the above embodiments of the present invention, there are at least the following beneficial effects: The present invention can integrate multi-dimensional data such as target electrocardiogram data, patient medical history information, location information, and symptom monitoring signals, and through the collaborative analysis of a chest pain symptom feature extraction model, a chest pain medical history feature extraction model, and a symptom signal feature extraction model, generate accurate diagnosis and treatment information including chest pain type, symptom description, and medical history record. This multi-modal data fusion method can improve the comprehensiveness and reliability of chest pain diagnosis and provide data support for clinical decision-making.

[0061] The present invention can dynamically update the diagnosis and treatment information database according to the chest pain diagnosis and treatment information, and generate chest pain trend information by analyzing the differences between the current chest pain distribution map and historical data, thereby optimizing the spatial configuration of first aid equipment. This intelligent analysis can improve the utilization efficiency of medical resources and provide a scientific basis for public health management and emergency response. Brief Description of the Drawings

[0062] By referring to the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understandable. In the drawings, several embodiments of the present invention are shown by way of example and not limitation, wherein:

[0063] Figure 1 is a schematic flowchart of a method for diagnosing and treating data analysis of chest pain patients provided by an embodiment of the present invention;

[0064] Figure 2 is a schematic structural diagram of a device for diagnosing and treating data analysis of chest pain patients provided by an embodiment of the present invention;

[0065] Figure 3 Schematically shows a schematic structural diagram of an electronic device according to an embodiment of the present invention. Detailed Embodiments

[0066] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and then implement the present invention, and not to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to be able to fully convey the scope of the present invention to those skilled in the art.

[0067] Those skilled in the art know that the embodiments of the present invention can be implemented as a device, apparatus, equipment, method, or computer program product. Therefore, the present invention can be specifically implemented in the following forms: completely hardware, completely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0068] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0069] The following refers to Figure 1 , Figure 1 is a schematic flowchart of a method for diagnosing and treating data analysis of chest pain patients provided by an embodiment of the present invention. As Figure 1 shown, a method for diagnosing and treating data analysis of chest pain patients includes:

[0070] S1 Obtain target electrocardiogram data, patient medical history information, location information, and target symptom signals, wherein the patient medical history information includes: past medical history records and medication history information, the location information represents the installation location of the medical device for collecting the target electrocardiogram data, and the target symptom signals are collected by a symptom monitoring device;

[0071] S2 uses a chest pain symptom feature extraction model to extract chest pain symptom features from the target electrocardiogram data, so as to generate chest pain symptom features, where the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features;

[0072] S3 uses a chest pain history feature extraction model to extract chest pain history features from the patient history information, so as to generate chest pain history features;

[0073] S4 uses a symptom signal feature extraction model to extract symptom signal features from the target symptom signal, so as to generate symptom signal features, where the chest pain symptom feature extraction model, the chest pain history feature extraction model, and the symptom signal feature extraction model are included in the chest pain diagnosis and treatment analysis model;

[0074] S5 determines chest pain diagnosis and treatment information according to the location information, the chest pain symptom features, the chest pain history features, and the symptom signal features, where the chest pain diagnosis and treatment information includes: location information, chest pain type, chest pain symptom description, and chest pain history record.

[0075] It should be noted that a method for diagnosing and treating data of chest pain patients proposed by the present invention aims to improve the accuracy and efficiency of chest pain diagnosis by comprehensively analyzing multiple data sources. Among them, the target electrocardiogram data refers to the electrocardiogram signals collected from patients, and these signals can reflect the electrical activity of the heart and are an important basis for diagnosing chest pain-related diseases. The patient history information includes past medical history records and medication history information. The past medical history record refers to the patient's past disease diagnosis and treatment experiences, and the medication history information involves the types, doses, and usage times of drugs that the patient has used. The location information refers to the geographical location of the medical device where the electrocardiogram data is collected, and this information helps to understand the location distribution of patients seeking medical treatment. The target symptom signal is collected by a symptom monitoring device, such as information on the pain intensity and duration during the patient's chest pain attack. By processing these data through a chest pain symptom feature extraction model, a chest pain history feature extraction model, and a symptom signal feature extraction model, chest pain diagnosis and treatment information including location information, chest pain type, chest pain symptom description, and chest pain history record is finally generated, providing a comprehensive diagnosis reference for doctors.

[0076] Specifically, the chest pain symptom feature extraction model is used to extract various features from the target electrocardiogram data. Among them, the electrocardiogram waveform features refer to the morphological, amplitude, and time interval features of different bands in the electrocardiogram (such as P wave, QRS complex, T wave, etc.), which can reflect the electrophysiological state of the heart. The pain location feature describes the specific location of the patient's chest pain, such as the front chest, back, or left chest. The pain nature feature includes the type of pain, such as a compressive sensation, stabbing pain, or burning sensation. The accompanying symptom feature involves whether the patient has other symptoms during the chest pain attack, such as sweating, difficulty breathing, or nausea. The attack time feature records the time pattern of the chest pain attack, such as whether it worsens after activity or occurs at night. The chest pain medical history feature extraction model extracts features from the patient's past medical history records and medication history information to analyze the patient's historical health status and drug use. The symptom signal feature extraction model processes the target symptom signal and extracts features related to the symptom. These models together constitute the chest pain diagnosis and treatment analysis model, which generates comprehensive chest pain diagnosis and treatment information through the comprehensive analysis of multi-source data.

[0077] Preferably, deep learning technology can be used to construct the chest pain symptom feature extraction model. For example, the basic feature generator can be a convolutional neural network (CNN) for extracting preliminary feature maps from the target electrocardiogram data. The complex feature encoder can further encode these feature maps to capture deeper information. The spatial anchor generator and feature region calibrator can be used to adjust and calibrate the feature maps spatially to ensure the accuracy of feature extraction. The first local feature extraction network set and the second local feature extraction network set can respectively perform local feature extraction on the spatial anchor feature map and the calibrated feature map to further refine the feature information. The electrocardiogram waveform feature extraction layer, pain location feature extraction layer, etc. can perform specialized extraction and processing for specific feature types. During the data processing, the electrocardiogram data can be preprocessed, such as filtering, denoising, etc. operations to improve the accuracy of feature extraction. At the same time, for the symptom signal feature extraction model, the target symptom signal can be normalized to better fuse and analyze with other features.

[0078] In some embodiments, the method further includes:

[0079] According to the chest pain type included in the chest pain diagnosis and treatment information, determine whether there is a diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information library;

[0080] In response to determining that there is no diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information library, add the chest pain diagnosis and treatment information as a diagnosis and treatment information record to the chest pain diagnosis and treatment information library;

[0081] In response to determining that there is a medical record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database, update the medical record corresponding to the chest pain diagnosis and treatment information according to the chest pain diagnosis and treatment information.

[0082] It should be noted that the present invention further expands the function of the chest pain diagnosis and treatment data analysis method. By comparing the generated chest pain diagnosis and treatment information with the chest pain diagnosis and treatment information database, dynamic update and management of medical records are realized. The chest pain diagnosis and treatment information database is a database that stores the medical treatment data of past chest pain patients. The medical records in it contain key information such as the chest pain type, symptom description, and medical history record of the patient. By judging whether there is a record corresponding to the current patient's chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database, the addition of new information or the update of existing information can be realized, so as to ensure the timeliness and accuracy of the medical record database. This process not only helps to accumulate clinical data, but also provides a more comprehensive reference for subsequent medical decisions.

[0083] Specifically, the construction of the chest pain diagnosis and treatment information database is to store and manage the medical treatment data of chest pain patients. After generating new chest pain diagnosis and treatment information, the system will judge whether there is a corresponding medical record in the diagnosis and treatment information database according to the key field of the chest pain type. The chest pain type refers to the possible cause type of chest pain comprehensively judged according to factors such as the patient's symptoms, electrocardiogram characteristics, and medical history, such as angina pectoris, myocardial infarction, pulmonary embolism, etc. If there is no record corresponding to the current chest pain diagnosis and treatment information in the diagnosis and treatment information database, it means that this is a new case type or new patient data, and it needs to be added to the diagnosis and treatment information database as a new medical record; if there is a corresponding record, the existing record needs to be updated according to the current chest pain diagnosis and treatment information, and the updated content may include new symptom descriptions, medical history supplements, or treatment suggestions, etc. This process can be realized through database query and update operations to ensure that the diagnosis and treatment information database can reflect the latest diagnosis and treatment data and knowledge in real time.

[0084] Preferably, to achieve efficient management and update of the chest pain diagnosis and treatment information database, a Structured Query Language (SQL) or a similar database management system tool can be used. When constructing the diagnosis and treatment information database, a unique identifier can be set for each diagnosis and treatment information record for quick location and update. When determining whether there is a corresponding diagnosis and treatment information record in the chest pain diagnosis and treatment information database, query conditions can be set, such as keyword matching of chest pain type, symptom description, etc., to quickly retrieve the database. If the retrieval result shows that there is no corresponding record, the system can automatically call the insert operation to add the new chest pain diagnosis and treatment information as a new record to the database. If there is a corresponding record, the key data fields (such as symptom description, medical history record, etc.) in the new chest pain diagnosis and treatment information can be updated to the existing record through the update operation. In addition, to ensure the accuracy and integrity of the data, a data verification and audit mechanism can be set when updating the diagnosis and treatment information record, such as through manual audit or automated rule checking, to ensure that the updated data meets the clinical medical standards and logic.

[0085] In some embodiments, the method further includes:

[0086] Read the updated chest pain diagnosis and treatment information database to obtain a set of diagnosis and treatment information records;

[0087] Determine the current chest pain distribution map according to the set of diagnosis and treatment information records;

[0088] Obtain the historical chest pain distribution map;

[0089] Generate chest pain trend information according to the current chest pain distribution map and the historical chest pain distribution map.

[0090] It should be noted that the present invention further expands the chest pain diagnosis and treatment data analysis method. By reading and analyzing the updated chest pain diagnosis and treatment information database, the current chest pain distribution map is generated, and the chest pain trend information is generated in combination with the historical chest pain distribution map. Here, the set of diagnosis and treatment information records refers to all the relevant data of chest pain patients read from the updated chest pain diagnosis and treatment information database, including information such as the chest pain type, symptom description, medical history record of the patients. The current chest pain distribution map is a visualization tool used to display the distribution of chest pain patients in different regions during the current time period to help with the rational allocation of medical resources. At the same time, the historical chest pain distribution map is a distribution map generated based on the data in the past time period and is used for comparison and analysis of the occurrence trend of chest pain. By comparing the current and historical chest pain distribution maps, chest pain trend information can be generated to provide data support for medical decision-making.

[0091] Specifically, the diagnosis and treatment information record set is a data set read from the chest pain diagnosis and treatment information database, which contains the chest pain diagnosis and treatment information of all patients. After being sorted and analyzed, this information is used to generate the current chest pain distribution map. The current chest pain distribution map can be a Geographic Information System (GIS) map, which marks the number and type distribution of chest pain patients in different regions. The historical chest pain distribution map is a similar map generated based on data over a past period of time for comparative analysis. The generation of chest pain trend information can be achieved by comparing the data in the current and historical distribution maps, analyzing the increasing or decreasing trends of chest pain in different regions, and the changes in different chest pain types. For example, if the number of chest pain patients in a certain region has been continuously increasing in the past few months, this may indicate potential health risks or increased demand for medical resources in that region.

[0092] Preferably, the process of generating the current chest pain distribution map and the historical chest pain distribution map can be realized with the aid of Geographic Information System (GIS) technology. First, extract data such as the geographical location information and chest pain type of each patient from the diagnosis and treatment information record set. Then, map this data onto the GIS map, and use different colors or icons to represent different types of chest pain and the number of patients. For the historical chest pain distribution map, a time range can be set, such as data for the past year or the past six months, and the map can be generated in the same way. When generating chest pain trend information, a detailed report or visualization chart can be generated by calculating the change rate of the number of chest pain patients in each region in the current and historical distribution maps, as well as the change trends of different chest pain types. For example, if the number of patients with acute myocardial infarction in a certain region has increased by 20% in the past year, this trend information can provide an important basis for the allocation of medical resources and the formulation of public health policies.

[0093] In some embodiments, the method further includes:

[0094] Mark the first-aid devices to be updated in location according to the current chest pain distribution map and the chest pain trend information to generate the marked first-aid device information, and obtain the marked first-aid device information set;

[0095] Optimize the set location of the first-aid devices corresponding to each piece of marked first-aid device information in the marked first-aid device information set.

[0096] It should be noted that the present invention further expands the method for analyzing chest pain diagnosis and treatment data. By combining the current chest pain distribution map and chest pain trend information, the setting positions of first aid devices are optimized. The set of post-tagged first aid device information here refers to the set of relevant information of the first aid devices whose positions need to be adjusted according to the chest pain distribution and trend analysis results. The optimization of the first aid device setting position means adjusting the layout of the first aid devices according to the spatial distribution and time trend of chest pain patients, so as to improve the first aid efficiency and resource utilization rate. This process helps to ensure that the first aid devices can more reasonably cover high-risk areas, reduce the first aid response time, and improve the accessibility of medical services.

[0097] Specifically, the current chest pain distribution map shows the distribution of chest pain patients in different regions during the current time period, while the chest pain trend information reflects the change trend of chest pain in different regions and at different times. By analyzing this information, it can be determined which regions have an increase or decrease in the number of chest pain patients, and which regions have changes in the types of chest pain. For example, if the number of chest pain patients in a certain region continues to increase and is mainly concentrated in severe disease types such as acute myocardial infarction, then this region may require more first aid device support. Each piece of post-tagged first aid device information in the set of post-tagged first aid device information contains key information such as the current position of the device, the new position recommended to be adjusted to, and the device type. The process of optimizing the first aid device setting position can be realized through geographic information system (GIS) technology. By combining the chest pain distribution map and trend information, an optimized layout plan for first aid devices is generated.

[0098] Preferably, the optimization of the first aid device setting position can be achieved through the following steps: First, use GIS technology to perform spatial analysis on the current chest pain distribution map and chest pain trend information to determine the regions where first aid devices need to be adjusted. For example, a threshold can be set to judge whether the number of chest pain patients in a certain region exceeds the average level, or whether the chest pain growth trend in a certain region is significant. Then, according to the analysis results, a set of post-tagged first aid device information is generated, and each piece of post-tagged first aid device information contains information such as the current position of the device and the new position recommended to be adjusted to. Next, optimization algorithms such as genetic algorithms or simulated annealing algorithms can be used to optimize the layout of first aid devices. The input parameters include the number, type, and current distribution positions of first aid devices, as well as the key data in the chest pain distribution map and trend information. The optimization goal is to maximize the coverage range of first aid devices while minimizing the first aid response time. Finally, according to the output results of the optimization algorithm, adjust the setting positions of first aid devices and update the first aid device layout map. This process can be executed regularly to ensure that the first aid device layout can adapt to the changes in chest pain distribution.

[0099] In some embodiments, the chest pain symptom feature extraction model includes: a basic feature generator, a complex feature encoder, a spatial anchor generator, a feature region calibrator, a first set of local feature extraction networks, a second set of local feature extraction networks, an electrocardiogram waveform feature extraction layer, a pain location feature extraction layer, a pain nature feature extraction layer, an accompanying symptom feature extraction layer, and an onset time feature extraction layer. The network structures of the spatial anchor generator and the feature region calibrator are the same; and by using the chest pain symptom feature extraction model to extract chest pain symptom features from the target electrocardiogram data to generate chest pain symptom features, including:

[0100] Through the basic feature generator, perform basic feature extraction on the target electrocardiogram data to generate a basic electrocardiogram feature map;

[0101] Through the complex feature encoder, perform complex feature encoding on the basic electrocardiogram feature map to generate a set of complex electrocardiogram feature maps;

[0102] Through the spatial anchor generator and the basic electrocardiogram feature map, generate a set of spatial anchor feature maps;

[0103] Through the feature region calibrator and the target complex electrocardiogram feature map, generate a set of calibrated feature maps, where the target complex electrocardiogram feature map is a complex electrocardiogram feature map in the set of complex electrocardiogram feature maps with a target ratio of the corresponding feature map size to the image size of the target electrocardiogram data;

[0104] Through the first set of local feature extraction networks, perform local feature extraction on the spatial anchor feature maps in the set of spatial anchor feature maps to generate a first local feature map, and obtain a set of first local feature maps;

[0105] Through the second set of local feature extraction networks, perform local feature extraction on the calibrated feature maps in the set of calibrated feature maps to generate a second local feature map, and obtain a set of second local feature maps;

[0106] Concatenate and fuse the set of first local feature maps, the set of second local feature maps, and the complex electrocardiogram feature maps in the set of complex electrocardiogram feature maps other than the target complex electrocardiogram feature map to obtain a concatenated and fused feature map;

[0107] Input the concatenated and fused feature map into the electrocardiogram waveform feature extraction layer to obtain the electrocardiogram waveform features;

[0108] Input the concatenated and fused feature map into the pain location feature extraction layer to obtain the pain location features;

[0109] Input the feature map after concatenation and fusion into the pain nature feature extraction layer to obtain the pain nature feature;

[0110] Input the feature map after concatenation and fusion into the accompanying symptom feature extraction layer to obtain the accompanying symptom feature;

[0111] Input the feature map after concatenation and fusion into the onset time feature extraction layer to obtain the onset time feature.

[0112] It should be noted that the present invention further elaborates in detail the structure and working principle of the chest pain symptom feature extraction model. This model extracts various chest pain-related features from the target electrocardiogram data through the collaborative work of multiple modules, including electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features. The extraction of these features is based on the multi-dimensional analysis of electrocardiogram data, aiming to provide more comprehensive and accurate information support for subsequent chest pain diagnosis. Each module in the model, such as the basic feature generator, complex feature encoder, spatial anchor generator, etc., undertakes specific processing tasks and jointly constitutes a complex feature extraction process.

[0113] Specifically, the chest pain symptom feature extraction model includes multiple key modules. The basic feature generator is used to extract preliminary feature maps from the target electrocardiogram data, and these feature maps contain the basic morphological information of the electrocardiogram. The complex feature encoder further encodes the basic feature maps to extract deeper features, such as the complex relationships between different bands in the electrocardiogram. The spatial anchor generator and the feature region calibrator are used to perform spatial adjustment on the feature maps to ensure the accuracy of feature extraction. The first local feature extraction network set and the second local feature extraction network set respectively perform local feature extraction on the spatial anchor feature map and the calibrated feature map to further refine the feature information. The electrocardiogram waveform feature extraction layer, the pain location feature extraction layer, etc. perform specialized extraction and processing for specific feature types. Among them, the target ratio refers to the ratio in the set of complex electrocardiogram feature maps that matches the image size of the target electrocardiogram data to ensure that the size of the feature map is suitable for subsequent processing. The collaborative work of these modules enables the model to extract key information from electrocardiogram data from different perspectives.

[0114] Preferably, deep learning techniques can be used to construct the chest pain symptom feature extraction model. For example, the basic feature generator can be a convolutional neural network (CNN), which takes the target electrocardiogram data as input and extracts preliminary feature maps through convolutional layers and pooling layers. The complex feature encoder can adopt an encoder-decoder structure to encode the basic feature maps and extract deeper feature information. The spatial anchor generator and the feature region calibrator can perform spatial calibration on the feature maps based on the attention mechanism to highlight the features of important regions. The first set of local feature extraction networks and the second set of local feature extraction networks can respectively adopt multiple convolutional layers and activation functions to extract local features from different feature maps. The electrocardiogram waveform feature extraction layer, the pain location feature extraction layer, etc. can design specialized network structures for specific feature types, such as outputting corresponding features through fully connected layers and classifiers. During the model training process, the labeled electrocardiogram data can be used as the training set, and the model parameters can be optimized through the backpropagation algorithm to improve the accuracy and robustness of feature extraction.

[0115] In some embodiments, the chest pain medical history feature extraction model includes: a first feature transformation layer, a medical history multi-dimensional feature extraction network, a second feature transformation layer, a medical history dynamic feature extraction network, a feature aggregation layer, a first feature mapping layer, and a second feature mapping layer; and the extraction of chest pain medical history features from the patient's medical history information through the chest pain medical history feature extraction model to generate chest pain medical history features includes:

[0116] Through the first feature transformation layer, perform feature transformation processing on the patient's medical history information to generate a first transformed medical history feature;

[0117] Through the medical history multi-dimensional feature extraction network, perform multi-dimensional feature extraction on the first transformed medical history feature to obtain multi-dimensional medical history features;

[0118] Through the second feature transformation layer, perform feature transformation processing on the multi-dimensional medical history features to generate a second transformed medical history feature;

[0119] Through the medical history dynamic feature extraction network, perform dynamic feature extraction on the second transformed medical history feature to obtain dynamic medical history features;

[0120] Through the feature aggregation layer, perform feature aggregation processing on the dynamic medical history features to generate aggregated dynamic medical history features;

[0121] Through the first feature mapping layer, perform a first feature mapping on the aggregated dynamic medical history features to obtain a first mapped dynamic medical history feature;

[0122] Through the second feature mapping layer, perform a secondary feature mapping on the dynamically mapped medical history features after the first mapping to obtain the chest pain medical history features.

[0123] It should be noted that the present invention further elaborates in detail on the structure and working principle of the chest pain medical history feature extraction model. This model generates chest pain medical history features that can reflect the patient's historical health status by performing multi-dimensional feature extraction and dynamic processing on the patient's medical history information. Each module in the model, such as the first feature transformation layer, the medical history multi-dimensional feature extraction network, the second feature transformation layer, the medical history dynamic feature extraction network, etc., undertakes specific processing tasks and jointly constitutes a complex feature extraction process. This process helps to make full use of the patient's medical history information and provides more comprehensive support for the diagnosis of chest pain.

[0124] Specifically, the chest pain medical history feature extraction model includes multiple key modules. The first feature transformation layer is used to perform preliminary feature transformation processing on the patient's medical history information, converting the original medical history data into a form suitable for subsequent processing. The medical history multi-dimensional feature extraction network is used to extract multi-dimensional features from the transformed medical history features, and these features may include multiple dimensions such as the patient's past disease types, incidence frequencies, treatment effects, etc. The second feature transformation layer further transforms the multi-dimensional medical history features to meet the requirements of subsequent dynamic feature extraction. The medical history dynamic feature extraction network is used to capture the dynamic changes in the medical history features, such as the patient's disease progression, changes in drug use, etc. The feature aggregation layer is used to perform aggregation processing on the dynamic medical history features to generate a more compact feature representation. The first feature mapping layer and the second feature mapping layer are used to further map and transform the aggregated dynamic medical history features, and finally generate the chest pain medical history features. The collaborative work of these modules enables the model to extract valuable features for chest pain diagnosis from the patient's medical history information.

[0125] Preferably, deep learning techniques can be used to construct the chest pain medical history feature extraction model. For example, the first feature transformation layer can be a simple fully-connected neural network, which takes the patient's medical history information as input and converts the medical history information into a preliminary feature representation through linear transformation and activation functions. The medical history multi-dimensional feature extraction network can adopt a convolutional neural network (CNN) or a recurrent neural network (RNN) structure to extract features according to the time series characteristics or spatial correlation of the medical history information. The second feature transformation layer can adopt a similar fully-connected layer structure to further transform the multi-dimensional medical history features. The medical history dynamic feature extraction network can adopt a long short-term memory network (LSTM) or a gated recurrent unit (GRU), and these network structures can effectively capture the dynamic changes in the medical history features. The feature aggregation layer can adopt a pooling layer or an attention mechanism to aggregate the dynamic medical history features. The first feature mapping layer and the second feature mapping layer can adopt a multi-layer perceptron (MLP) structure to further map and transform the aggregated features. During the model training process, the labeled medical history data can be used as the training set, and the model parameters can be optimized through the backpropagation algorithm to improve the accuracy and robustness of feature extraction.

[0126] In some embodiments, the medication history information includes: a set of medication component description information, and the medication component information in the set of medication component information includes: medication component type and dosage; the symptom signal feature extraction model includes: a symptom multi-dimensional feature extractor and a feature fusion mapping layer; and the extracting of symptom signal features from the target symptom signal through the symptom signal feature extraction model to generate symptom signal features includes:

[0127] Through the symptom multi-dimensional feature extractor, multi-dimensional feature extraction is performed on each piece of medication component information in the set of medication component information to generate symptom component features, and a set of symptom component features is obtained;

[0128] Feature fusion is performed on the symptom component features in the set of symptom component features to obtain the fused symptom component features;

[0129] Range mapping processing is performed on the symptom signal to generate the range-mapped symptom signal;

[0130] The range-mapped symptom signal and the fused symptom component features are feature concatenated to obtain the concatenated features;

[0131] The concatenated features are input into the feature fusion mapping layer to obtain the symptom signal features.

[0132] It should be noted that the present invention further elaborates in detail the structure and working principle of the symptom signal feature extraction model. This model generates symptom signal features that can reflect the current symptom state of the patient by performing multi-dimensional feature extraction and fusion processing on the target symptom signal. Each module in the model, such as the symptom multi-dimensional feature extractor and the feature fusion mapping layer, undertakes specific processing tasks and together constitutes a complex feature extraction process. This process helps to make full use of the symptom signal information of the patient and provides more comprehensive support for the diagnosis of chest pain.

[0133] Specifically, the symptom signal feature extraction model includes two key modules: the symptom multi-dimensional feature extractor and the feature fusion mapping layer. The symptom multi-dimensional feature extractor is used to perform multi-dimensional feature extraction on each piece of medication component information in the medication component information set, generating a set of symptom component features. Each piece of medication component information in the medication component information set contains the type and dosage of the medication component, and these information reflect the current medication situation of the patient. The symptom multi-dimensional feature extractor extracts features related to symptoms by analyzing this information. The feature fusion mapping layer is used to further process the fused symptom component features and the symptom signal after range mapping to generate the final symptom signal features. The range mapping process refers to adjusting the numerical range of the symptom signal to a suitable interval for subsequent feature fusion and analysis. The collaborative work of these modules enables the model to extract valuable features for chest pain diagnosis from the patient's symptom signals.

[0134] Preferably, deep learning techniques can be used to construct the symptom signal feature extraction model. For example, the symptom multi-dimensional feature extractor can be a multi-layer perceptron (MLP) or a convolutional neural network (CNN), whose input is each piece of medication component information in the medication component information set, and multi-dimensional features related to symptoms are extracted through a multi-layer neural network. The feature fusion mapping layer can adopt a fully connected neural network, whose input is the fused symptom component features and the symptom signal after range mapping, and the final symptom signal features are generated through linear transformation and activation functions. During the model training process, labeled symptom signal data can be used as the training set, and the model parameters can be optimized through the backpropagation algorithm to improve the accuracy and robustness of feature extraction. For example, the patient's symptom signals and the corresponding diagnosis results can be used as training samples to train the model to learn the mapping relationship between the symptom signals and the types of chest pain. In addition, to improve the generalization ability of the model, regularization techniques such as weight decay or Dropout can be introduced during the training process to prevent the model from overfitting.

[0135] The above embodiments of the present invention have the following beneficial effects: The present invention can realize the intelligent diagnosis and analysis of chest pain patients based on multi-source medical data. By integrating electrocardiogram data, medical history records, location information, and real-time symptom monitoring data, a comprehensive chest pain diagnosis and treatment information can be constructed. Using a feature extraction model to deeply analyze electrocardiogram waveforms, pain characteristics, medical history data, and medication information, a structured diagnosis and treatment report including chest pain type, symptom description, and medical history record can be generated, providing an accurate basis for clinical decision-making. By establishing a chest pain diagnosis and treatment information database and implementing a dynamic update mechanism, the diagnosis model can be continuously optimized and clinical experience data can be accumulated.

[0136] Based on the spatio-temporal distribution characteristics of chest pain diagnosis and treatment information, the present invention can realize the intelligent allocation and optimization of medical resources. By comparing and analyzing the differences between the current chest pain distribution and historical data, chest pain trend prediction information can be generated, providing decision-making support for public health management. At the same time, by marking the first-aid equipment to be updated and optimizing its spatial configuration, the first-aid response efficiency can be improved. This technical solution can realize a closed-loop service from individual diagnosis and treatment to population health management, improving both the clinical diagnosis accuracy and the efficiency of medical resource allocation.

[0137] As Figure 2 shown, a diagnostic data analysis device for chest pain patients in some embodiments, the device includes:

[0138] An acquisition unit 201, configured to acquire target electrocardiogram data, patient medical history information, location information, and target symptom signals, where the patient medical history information includes: past medical history records and medication history information, the location information represents the installation location of the medical device that collects the target electrocardiogram data, and the target symptom signals are collected by a symptom monitoring device;

[0139] A chest pain symptom feature extraction unit 202, configured to extract chest pain symptom features from the target electrocardiogram data through a chest pain symptom feature extraction model to generate chest pain symptom features, where the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features;

[0140] A chest pain medical history feature extraction unit 203, configured to extract chest pain medical history features from the patient medical history information through a chest pain medical history feature extraction model to generate chest pain medical history features;

[0141] A symptom signal feature extraction unit 204, configured to extract symptom signal features from the target symptom signals through a symptom signal feature extraction model to generate symptom signal features, where the chest pain symptom feature extraction model, the chest pain medical history feature extraction model, and the symptom signal feature extraction model are included in the chest pain diagnosis and treatment analysis model;

[0142] A determination unit 205, configured to determine chest pain diagnosis and treatment information according to the position information, the chest pain symptom characteristics, the chest pain medical history characteristics, and the symptom signal characteristics, where the chest pain diagnosis and treatment information includes: position information, chest pain type, chest pain symptom description, and chest pain medical history record.

[0143] It can be understood that the various modules described in the chest pain patient's diagnosis and treatment data analysis device correspond to the respective steps in the chest pain patient's diagnosis and treatment data analysis method described in the reference Figure 1 Therefore, the operations, features, and beneficial effects described above for the chest pain patient's diagnosis and treatment data analysis method also apply to the chest pain patient's diagnosis and treatment data analysis device and the modules included therein, and will not be elaborated herein.

[0144] Next, refer to Figure 3 , which shows a schematic structural diagram of an electronic device 300 suitable for implementing some embodiments of the present invention. The electronic device in some embodiments of the present invention may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The terminal device shown is only an example and should not impose any limitations on the functions and usage scopes of the embodiments of the present invention.

[0145] As Figure 3 shown, the electronic device 300 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0146] Generally, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 3An electronic device 300 having various devices is shown, but it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 3 Each block shown in Figure 3 may represent a device or, as needed, multiple devices.

[0147] Furthermore, the storage medium of the embodiments of the present application stores program instructions capable of implementing all the above methods. Among them, the program instructions may be stored in the above storage medium in the form of a software product, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: various media capable of storing program codes such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc, or a terminal device such as a computer, a server, a mobile phone, or a tablet.

[0148] The above description is only some preferred embodiments of the present invention and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, a technical solution formed by mutually replacing the above features with the (but not limited to) technical features having similar functions disclosed in the embodiments of the present invention.

Claims

1. A method for analyzing diagnosis and treatment data of patients with chest pain, characterized in that: include: Acquire target electrocardiogram data, patient medical history information, location information and target symptom signal, wherein the patient medical history information includes: past medical history records and medication history information, the location information represents the setting location of the medical device that collects the target electrocardiogram data, and the target symptom signal is collected by a symptom monitoring device; Extracting chest pain symptom features from the target electrocardiogram data through a chest pain symptom feature extraction model to generate chest pain symptom features, wherein the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features; Extracting chest pain history features from the patient's medical history information using a chest pain history feature extraction model to generate chest pain history features; Performing symptom signal feature extraction on the target symptom signal through a symptom signal feature extraction model to generate symptom signal features, wherein the chest pain symptom feature extraction model, the chest pain history feature extraction model, and the symptom signal feature extraction model are included in a chest pain diagnosis and treatment analysis model; Chest pain diagnosis and treatment information is determined based on the location information, the chest pain symptom characteristics, the chest pain medical history characteristics and the symptom signal characteristics, wherein the chest pain diagnosis and treatment information includes: location information, chest pain type, chest pain symptom description and chest pain medical history record.

2. The method according to claim 1, characterized in that: The method further comprises: According to the chest pain type included in the chest pain diagnosis and treatment information, determining whether there is a diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database; In response to determining that there is no diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information in the chest pain diagnosis and treatment information database, adding the chest pain diagnosis and treatment information as a diagnosis and treatment information record to the chest pain diagnosis and treatment information database; In response to determining that a diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information exists in the chest pain diagnosis and treatment information database, the diagnosis and treatment information record corresponding to the chest pain diagnosis and treatment information is updated according to the chest pain diagnosis and treatment information.

3. The method according to claim 2, characterized in that The method further comprises: Read the updated chest pain diagnosis and treatment information database to obtain a diagnosis and treatment information record set; Determine a current chest pain distribution map according to the diagnosis and treatment information record set; Obtain a historical chest pain profile; Chest pain trend information is generated according to the current chest pain distribution map and the historical chest pain distribution map.

4. The method according to claim 3, characterized in that The method further comprises: According to the current chest pain distribution map and the chest pain trend information, marking the emergency equipment to be updated in position to generate marked emergency equipment information, and obtaining a marked emergency equipment information set; The first aid equipment setting position is optimized for the first aid equipment corresponding to each piece of marked first aid equipment information in the marked first aid equipment information set.

5. The method according to claim 4, characterized in that The chest pain symptom feature extraction model includes: a basic feature generator, a complex feature encoder, a spatial anchor generator, a feature region calibrator, a first local feature extraction network set, a second local feature extraction network set, an electrocardiogram waveform feature extraction layer, a pain location feature extraction layer, a pain property feature extraction layer, an accompanying symptom feature extraction layer and an onset time feature extraction layer, and the network structure of the spatial anchor generator is consistent with that of the feature region calibrator; and the chest pain symptom feature extraction model is used to extract chest pain symptom features from the target electrocardiogram data to generate chest pain symptom features, including: Extracting basic features from the target electrocardiogram data by means of the basic feature generator to generate a basic electrocardiogram feature graph; Performing complex feature encoding on the basic electrocardiogram feature graph by the complex feature encoder to generate a complex electrocardiogram feature graph set; Generate a set of spatial anchor point feature maps by using the spatial anchor point generator and the basic electrocardiogram feature map; Generate a calibrated feature map set by using the feature region calibrator and the target complex electrocardiogram feature map, wherein the target complex electrocardiogram feature map is a complex electrocardiogram feature map in the complex electrocardiogram feature map set, the corresponding feature map size of which is a target ratio of the image size of the target electrocardiogram data; Performing local feature extraction on the spatial anchor feature maps in the spatial anchor feature map set through a first local feature extraction network set to generate a first local feature map, thereby obtaining a first local feature map set; Performing local feature extraction on the calibrated feature maps in the calibrated feature map set by using the second local feature extraction network set to generate second local feature maps, thereby obtaining a second local feature map set; The first local feature map set, the second local feature map set, and the complex electrocardiogram feature map set except the target complex electrocardiogram feature map are fused in series to obtain a fused feature map; Inputting the serially fused feature graph into the electrocardiogram waveform feature extraction layer to obtain the electrocardiogram waveform feature; Inputting the serially fused feature map into the pain site feature extraction layer to obtain the pain site feature; Inputting the serially fused feature map into the pain property feature extraction layer to obtain the pain property feature; Inputting the serially fused feature graph into the accompanying symptom feature extraction layer to obtain the accompanying symptom feature; The serially fused feature map is input into the onset time feature extraction layer to obtain the onset time feature.

6. The method according to claim 5, characterized in that The chest pain history feature extraction model includes: a first feature transformation layer, a medical history multidimensional feature extraction network, a second feature transformation layer, a medical history dynamic feature extraction network, a feature aggregation layer, a first feature mapping layer and a second feature mapping layer; and the chest pain history feature extraction model is used to extract chest pain history features from the patient's medical history information to generate chest pain history features, including: Performing feature transformation processing on the patient's medical history information through the first feature transformation layer to generate a first transformed medical history feature; Performing multidimensional feature extraction on the first transformed medical history features through a medical history multidimensional feature extraction network to obtain multidimensional medical history features; Performing feature transformation processing on the multidimensional medical history feature through the second feature transformation layer to generate a second transformed medical history feature; Performing dynamic feature extraction on the second transformed medical history feature through a medical history dynamic feature extraction network to obtain a dynamic medical history feature; Performing feature aggregation processing on the dynamic medical history features through the feature aggregation layer to generate aggregated dynamic medical history features; Performing a feature mapping on the dynamic medical history features after the aggregation process through the first feature mapping layer to obtain a dynamic medical history feature after the first mapping; The second feature mapping layer performs secondary feature mapping on the dynamic medical history features after the primary mapping to obtain the chest pain medical history features.

7. The method according to claim 6, characterized in that The medication history information includes: a medication component description information set, the medication component information in the medication component information set includes: medication component type and medication dosage, the symptom signal feature extraction model includes: a symptom multidimensional feature extractor and a feature fusion mapping layer; and the symptom signal feature extraction model is used to extract the symptom signal feature of the target symptom signal to generate a symptom signal feature, including: By means of the symptom multidimensional feature extractor, multidimensional feature extraction is performed on each medication component information in the medication component information set to generate symptom component features, thereby obtaining a symptom component feature set; Performing feature fusion on the symptom component features in the symptom component feature set to obtain fused symptom component features; performing range mapping processing on the symptom signal to generate a range-mapped symptom signal; Performing feature splicing on the range-mapped symptom signal and the fused symptom component feature to obtain a spliced ​​feature; The spliced ​​features are input into the feature fusion mapping layer to obtain the symptom signal features.

8. A diagnostic data analysis device for chest pain patients, characterized in that: include: an acquisition unit configured to acquire target electrocardiogram data, patient medical history information, location information, and target symptom signals, wherein the patient medical history information includes: past medical history records and medication history information, the location information represents the setting location of the medical device for collecting the target electrocardiogram data, and the target symptom signals are collected by a symptom monitoring device; a chest pain symptom feature extraction unit, configured to extract chest pain symptom features from the target electrocardiogram data through a chest pain symptom feature extraction model to generate chest pain symptom features, wherein the chest pain symptom features include: electrocardiogram waveform features, pain location features, pain nature features, accompanying symptom features, and onset time features; a chest pain medical history feature extraction unit, configured to extract chest pain medical history features from the patient medical history information through a chest pain medical history feature extraction model to generate chest pain medical history features; a symptom signal feature extraction unit, configured to extract symptom signal features from the target symptom signal through a symptom signal feature extraction model to generate symptom signal features, wherein the chest pain symptom feature extraction model, the chest pain history feature extraction model, and the symptom signal feature extraction model are included in a chest pain diagnosis and treatment analysis model; The determination unit is configured to determine chest pain diagnosis and treatment information based on the location information, the chest pain symptom characteristics, the chest pain medical history characteristics and the symptom signal characteristics, wherein the chest pain diagnosis and treatment information includes: location information, chest pain type, chest pain symptom description and chest pain medical history record.

9. An electronic device, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer-readable medium, characterized in that A computer program is stored thereon, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images