阅读说明：本技术 一种流行病患者信息三维空间可视化方法 (Three-dimensional space visualization method for information of epidemic patient ) 是由 江南 陈云海 张帆 曹一冰 张江水 崔虎平 张政 陈敏颉 于 2020-12-10 设计创作，主要内容包括：本发明涉及一种流行病患者信息三维空间可视化方法,属于疾病地理学技术领域。本发明首先对获取的患者信息进行多粒度时空对象的构建,以多粒度时空对象的形式对患者信息进行描述和分析；然后将组织成多粒度时空对象的患者信息加载到具有三维空间的基础场景中,实现三维动态可视化表达；最后根据需要进行不同层次的可视分析。本发明结合多粒度时空对象模型对患者信息进行可视化的方法更加直观,能够帮助用户快速发现其潜在传播规律和进行疫情风险评估,且可移植性好、拓展性能佳。(The invention relates to a three-dimensional space visualization method for information of epidemic patients, and belongs to the technical field of disease geography. Firstly, constructing a multi-granularity space-time object on the acquired patient information, and describing and analyzing the patient information in the form of the multi-granularity space-time object; then loading the patient information organized into multi-granularity space-time objects into a basic scene with a three-dimensional space to realize three-dimensional dynamic visual expression; and finally, performing visual analysis of different layers according to the requirement. The method for visualizing the patient information by combining the multi-granularity space-time object model is more intuitive, can help the user to quickly find the potential propagation rule and evaluate the epidemic situation risk, and has good transportability and good expansion performance.)

1. A three-dimensional space visualization method for information of popular patients is characterized by comprising the following steps:

1) acquiring patient information, and preprocessing the patient information to obtain corresponding basic information, infection relation and dynamic characteristics;

2) realizing basic information modeling, infection relation modeling and dynamic characteristic modeling based on a multi-granularity space-time object model, and organizing basic information, infection relation and dynamic characteristic information of a patient into a multi-granularity space-time object;

3) building a basic scene by using a digital earth framework, loading patient information organized into multi-granularity space-time objects into the basic scene, and displaying the patient information according to visual view design content;

4) and performing visual analysis on the displayed content according to actual requirements to obtain visual results of different layers.

2. The method for visualizing the three-dimensional space of the information of popular patients as claimed in claim 1, wherein the patient information in the step 1) is obtained by means of Python tool crawling.

3. The three-dimensional space visualization method for popular patient information according to claim 1, wherein the multi-granularity spatiotemporal objects organized in step 2) are stored by using a mashup database, which comprises a master database and a slave database, wherein the master database is used for storing patient basic information, and the slave database is used for storing patient infection relations and dynamic characteristics, respectively.

4. The three-dimensional space visualization method for information about pandemic patients as claimed in claim 1, wherein the visualization view design content includes a patient management and interactive view, a basic information display view, a scene display and interactive view and a time management and backtrack view; the patient management and interactive view is used for realizing interactive loading, management and list display of the objects; the basic information display view is used for displaying the object multi-dimensional basic characteristic information; the scene display and interactive view is a main view of patient information visualization and is used for visually displaying the spatial form, the spatial position, the spatial relation and the space-time dynamic process of a patient from a geographical view; the time management and backtracking view is used for realizing the control of the system simulation time and the backtracking of the object space-time process in an interactive mode.

5. The method for three-dimensional space visualization of pandemic patient information according to claim 1 or 4, wherein the visual analysis in step 4) comprises at least one of a macro-level visual analysis, a meso-level visual analysis and a micro-level visual analysis;

the macroscopic layer visual analysis is used for establishing a spatial-temporal evolution process of the patient by adopting an equivalent region method and a partition statistical chart and dynamically expressing the characteristic information of the patient;

the mesoscopic visual analysis is used for dynamically expressing the characteristic information of the patient in the form of a simulated object track spatiotemporal process, a spatial form spatiotemporal process and an attribute spatiotemporal process;

the microscopic visual analysis refers to deep mining and visual analysis of patient information by adopting a method for simulating the time-space change process of a patient infection link and analyzing hot spots of an easily infected area.

6. The method for three-dimensional space visualization of pandemic patient information according to claim 5, wherein the trajectory spatiotemporal process describes a spatial movement process of the patient for a period of time before the patient is diagnosed; the space-time process of the spatial morphology expresses the change process of the icon morphology in the process of patient before diagnosis, suspected diagnosis, confirmed diagnosis, cure/death; the attribute spatiotemporal process characterizes the dynamic change process of the basic characteristic information of the patient.

7. The method of three-dimensional space visualization of pandemic patient information as claimed in claim 5 wherein the infection link is a time series directed graph with nodes representing patients and node relationships connected by three-dimensional directed arcs.

8. The epidemic patient information three-dimensional space visualization method according to claim 5, wherein the susceptibility area hot spot analysis displays the hot spot areas of the patient staying in a period of time before diagnosis in a dynamic thermodynamic diagram manner, and the susceptibility degree of the susceptibility area is differentiated by the difference and shade of colors.

Technical Field

The invention relates to a three-dimensional space visualization method for information of epidemic patients, and belongs to the technical field of disease geography.

Background

Visualization is an important means for assisting human beings to discover implicit rules of data, and is also an effective way for transmitting complex information. In 1854, UK doctor John Snow discovered the source of the outbreak of cholera in the Brad district of London by using a map visualization method according to the information of patients. Then, a plurality of scholars continuously develop deep research and experiments aiming at contents such as map expression, visual system design, space analysis and the like of epidemic disease information, generate various map works, visual systems and analysis tools, can visually express and analyze the time-space difference condition of the epidemic disease patient information, and timely discover potential epidemic disease rules and knowledge. In particular, since the prevalence of new coronavirus pneumonia (coronavirus disease 2019, COVID-19), researchers have reported many representative works around COVID-19 visualization, such as Mocnik et al visually analyzing the COVID-19 epidemic situation in China, Europe, and the United states by combining thematic map, polyline statistical map, and temporal band diagram. For example, an interactive visual epidemic situation analysis method using a geographical knowledge map is proposed, which can realize visual analysis of the epidemic situation, but focuses on modeling and analysis of macro information such as the number of national and urban patients, does not comprehensively and systematically depict the epidemic situation from multiple granularities and multiple levels, describes and expresses patient information in an all-round manner, and has poor visualization effect.

Disclosure of Invention

The invention aims to provide a three-dimensional space visualization method for information of popular patients, which aims to solve the problem of poor visualization effect in the current visualization process.

The invention provides a three-dimensional space visualization method for popular patient information to solve the technical problems, which comprises the following steps:

1) acquiring patient information, and preprocessing the patient information to obtain corresponding basic information, infection relation and dynamic characteristics;

2) realizing basic information modeling, infection relation modeling and dynamic characteristic modeling based on a multi-granularity space-time object model, and organizing basic information, infection relation and dynamic characteristic information of a patient into a multi-granularity space-time object;

3) building a basic scene by using a digital earth framework, loading patient information organized into multi-granularity space-time objects into the basic scene, and displaying the patient information according to visual view design content;

4) and performing visual analysis on the displayed content according to actual requirements to obtain visual results of different layers.

Firstly, constructing a multi-granularity space-time object on the acquired patient information, and describing and analyzing the patient information in the form of the multi-granularity space-time object; then loading the patient information organized into multi-granularity space-time objects into a basic scene with a three-dimensional space to realize three-dimensional dynamic visual expression; and finally, performing visual analysis of different layers according to the requirement. The method for visualizing the patient information by combining the multi-granularity space-time object model is more intuitive, can help the user to quickly find the potential propagation rule and evaluate the epidemic situation risk, and has good transportability and good expansion performance.

Further, in order to accurately and conveniently acquire the patient information, the patient information in the step 1) is obtained by crawling through a Python tool.

Further, in order to conveniently realize the scheduling of the multi-granularity data, the multi-granularity spatiotemporal objects organized in the step 2) are stored by adopting a mixed database, which comprises a master database and a slave database, wherein the master database is used for storing the basic information of the patient, and the slave database is used for respectively storing the infection relation and the dynamic characteristics of the patient.

Further, in order to realize comprehensive display, the visual view design content comprises a patient management and interaction view, a basic information display view, a scene display and interaction view and a time management and backtracking view; the patient management and interactive view is used for realizing interactive loading, management and list display of the objects; the basic information display view is used for displaying the object multi-dimensional basic characteristic information; the scene display and interactive view is a main view of patient information visualization and is used for visually displaying the spatial form, the spatial position, the spatial relation and the space-time dynamic process of a patient from a geographical view; the time management and backtracking view is used for realizing the control of the system simulation time and the backtracking of the object space-time process in an interactive mode.

Further, in order to realize the display of different levels and meet the requirement of diversity, the visual analysis in the step 4) comprises at least one of macroscopic level visual analysis, mesoscopic level visual analysis and microscopic level visual analysis;

the macroscopic layer visual analysis is used for establishing a spatial-temporal evolution process of the patient by adopting an equivalent region method and a partition statistical chart and dynamically expressing the characteristic information of the patient;

the mesoscopic visual analysis is used for dynamically expressing the characteristic information of the patient in the form of a simulated object track spatiotemporal process, a spatial form spatiotemporal process and an attribute spatiotemporal process;

the microscopic visual analysis refers to deep mining and visual analysis of patient information by adopting a method for simulating the time-space change process of a patient infection link and analyzing hot spots of an easily infected area.

Further, the trajectory spatiotemporal process describes a spatial movement process of the patient for a period of time before the patient is diagnosed; the space-time process of the spatial morphology expresses the change process of the icon morphology in the process of patient before diagnosis, suspected diagnosis, confirmed diagnosis, cure/death; the attribute spatiotemporal process characterizes the dynamic change process of the basic characteristic information of the patient.

Further, for convenient and accurate description of the infection link chain, the infection link chain is a time sequence directed graph, wherein nodes represent patients, and the node relations are connected through three-dimensional directed radians.

Further, in order to more intuitively know the susceptible area, the hot spot analysis of the susceptible area displays the hot spot area remained in a period of time before the patient is diagnosed in a dynamic thermodynamic diagram mode, and the susceptible degree of the susceptible area is distinguished by using different colors and different shades.

Drawings

FIG. 1 is a flow chart of a three-dimensional space visualization method of popular patient information according to the present invention;

FIG. 2 is a schematic diagram of a process for modeling suspected infection relationships of a patient according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of modeling a spatiotemporal process of patient dynamic characteristics in an embodiment of the invention;

FIG. 4 is a flow chart of patient information objectification processing in an embodiment of the present invention;

FIG. 5 is a flow chart of a visualization presentation of patient information in an embodiment of the present invention;

FIG. 6 is a diagram illustrating the visual analysis of patient information at a mid-view level in an embodiment of the present invention;

FIG. 7 is a diagram of the effect of visual analysis of microscopic patient information in an embodiment of the present invention.

Detailed Description

The following further describes embodiments of the present invention with reference to the drawings.

The space-time object abstracts and describes the real world in a manner easy for human cognition, and can embody the space-time characteristics and dynamic characteristics of the space entity from different levels and different granularities. The space-time object is based on a space-time data model, the time dimension of the space-time object is mainly expanded, and the space-time object is a data basis for space data analysis and visual expression. With the development of the technology, the current space-time object model is developed into a multi-granularity space-time object model, and compared with the traditional space-time object model, the model has the characteristics of multi-granularity, multi-type, multi-form, multi-reference system, multi-element association, multi-dimensional dynamics and the like, and can comprehensively describe and express multi-element characteristic information of the object. The invention combines the multi-granularity time-space object modeling idea with the time-space information visualization technical method, performs visual analysis on the basis of constructing the patient time-space object, can fill the blank of the research of the multi-granularity and multi-level patient information visual analysis method to a certain extent, and provides a more intuitive analysis method for epidemic prevention and control.

As shown in figure 1, the invention firstly obtains the information of the patient, and obtains the basic information, the infection relation and the dynamic characteristics through preprocessing; then realizing basic information modeling, infection relation modeling and dynamic characteristic modeling based on a multi-granularity space-time object model, and organizing basic information, infection relation and dynamic characteristic information of a patient into a multi-granularity space-time object; building a basic scene by using a digital earth framework, loading the patient information organized into multi-granularity space-time objects into the basic scene, and displaying according to the designed visual view content; and performing visual analysis on the displayed content according to actual requirements to obtain visual results of different layers. The following describes the specific implementation process of the present invention in detail with reference to specific examples, which are as follows.

1. And acquiring the information of the patient, and preprocessing the information to obtain corresponding basic information, infection relation and dynamic characteristics.

The patient information refers to epidemic patient information (abbreviated as case information), is patient epidemiological survey data, mainly comes from the websites of health care committee and government officials of each province, and takes the information of the COVID-19 case as an example, and the main contents include basic information, a diagnosis confirming place, a behavior track and the like of the patient.

The invention can utilize Python tool to capture the patient information from the network and store it in the local database. Since there may be many interferences in the captured data, the data needs to be cleaned by writing an application program and manually checking, and interference information, such as noise and default values, needs to be removed. Because the captured patient information has too much content and is not strong in regularity, structured processing and infection relation reasoning are carried out on the patient information by utilizing a jieba word segmentation tool of Python, a Baidu map API and a scene semantic analysis method, wherein the word segmentation tool is mainly used for extracting spatio-temporal information, such as time and place in a behavior track text; the Baidu map API is mainly used for acquiring longitude and latitude of place names; the contextual semantic analysis is mainly used for generating the patient space-time trajectory series and reasoning the patient infection relation. Meanwhile, other information such as body temperature change data of the patient can be added according to actual display requirements.

2. Patient information is organized into multi-granular spatiotemporal objects based on a multi-granular spatiotemporal object model.

The spatiotemporal object construction aims at organizing patient information into multi-granularity spatiotemporal objects, and the modeling process can be divided into four parts, namely basic information modeling, infection relation modeling, dynamic characteristic modeling and objectification processing.

1) And modeling basic information.

The basic information is mainly used for describing basic characteristics of the patient, and the content of the basic information comprises an ID number, a time reference, a space position, a space form and the like of the patient, wherein the ID number is mainly used for uniquely identifying the patient, the time reference and the space reference are mainly used for providing reference standards for describing spatio-temporal characteristic information of the patient, the space position is mainly used for spatial positioning of the patient, and the space form is mainly used for representing a spatial presentation pattern of the patient. The method organizes object basic information into an object feature set by establishing rules and constraint conditions, namely OBInfo { OID, SRS, TRS, Position, Attributes, spatialForm and DateTime }, wherein the OBInfo is the object basic information, the OID is an object unique identifier, the SRS is a spatial reference set, the TRS is a time reference set, the Position is an object initial Position, the Attributes is an object attribute feature set, the spatialForm is an object initial spatial form, and DateTime is an object life cycle. Common modeling rules include multi-granularity space-time object formalization description, OID unique index constraint, object feature transformation rules and other constraint conditions, such as mapping relations of different space-time references and action ranges thereof, life cycle lengths, attribute feature value ranges and the like.

2) And (5) modeling infection relation.

The infection relation records information such as infection sources, infection objects, infection time and the like, and the information is expressed and managed in a whole life cycle by adopting an object-oriented mode. Patient infection is a one-to-many relationship, a collection of relationships. With "[ ]" representing a set and "{ }" representing a relationship, a formal description of a patient's infectious relationship can be expressed as: r [ { SO1, DO1, relationship type, StartTime, EndTime }, { SO1, DO2, relationship type, StartTime, EndTime }, …, { SO1, DOn, relationship type, StartTime, EndTime } ], where SO1 is an infection source patient, DOn is an infection target, relationship type is a relationship class, StartTime is a relationship creation time, and EndTime is a relationship end time. Patient infection relationships can be created, stored, and managed using a spatiotemporal series snapshot model, as shown in fig. 2.

3) And modeling dynamic characteristics.

The patient information includes a large amount of continuously changing dynamic characteristic information, such as a patient space-time trajectory, dynamic attributes, state information and the like, besides the basic information and the infection relation information, wherein the dynamic attributes mainly refer to continuously changing attribute information of the patient, such as attributes of the patient, such as the body temperature, the heart rate, the blood pressure and the like; the status information is mainly used for describing the health status of the patient, such as diagnosis or cure, and in combination with the disease process, the status information can describe the health status change process of the patient from pre-diagnosis-suspected-confirmed-cure or death, as shown in fig. 3. The dynamic characteristic attribute of the patient is an object characteristic set, and the formal description of the dynamic characteristic attribute is as follows: DOA [ { OID, TableName, DateTime1, Field1, Field2, …, Fieldn }, { OID, TableName, DateTime2, Field1, Field2, …, Fieldn }, …, { OID, TableName, DateTime, Field1, Field2, …, Fieldn } ], wherein DOA is a patient dynamic feature attribute, OID is an object unique identifier, TableName is an object dynamic feature information storage location, DateTime1-m is a dynamic feature time marker for indicating a time corresponding to a dynamic feature, such as a time at which a patient is in a respective state or location, Field1-n is a dynamic feature formalized description, such as a location, state information, and dynamic attributes of the patient. Different types of information are stored in different storage locations and may be distinguished using different databases.

4) And (5) carrying out objectification processing.

The essence of the objectification processing is the process of organizing basic information, infection relation and dynamic characteristic information of a patient into multi-granularity space-time objects under the constraint of a space-time domain and a class template, namely instantiating the space-time objects of the patient. And finally, storing the objectification result in a spatiotemporal object database or generating a corresponding data exchange format file (as shown in figure 4). The space-time object database mainly realizes the dynamic storage and access scheduling of multi-granularity space-time objects, and the exchange format is mainly used for object data sharing and transmission. The multi-granularity space-time object is stored by adopting a mixed database, wherein a main database (Postgresql) mainly stores basic information of patients, and contents such as infection relations, dynamic characteristics, space forms, data files and the like of the patients are respectively stored in slave databases Neo4j, Geomesa, Mongodb and HDFS. The exchange format is a file representation mode of the multi-granularity space-time object and is in import and export relation with the space-time object database. Common interchange formats are 3 forms of XML, JSON and binary files.

3. And (3) building a basic scene by using a digital earth framework, loading the patient information organized into the multi-granularity space-time object into the basic scene, and displaying according to the designed visual view content.

1) And determining a visual environment and a tool, and building a basic scene.

The patient information visualization adopts a C/S architecture, a VS2010 compiling development environment, a C + + programming language, a QT4.8.6x64 image interface framework, an osgEarth three-dimensional digital earth framework and a Google Protocol Buffer network transmission sequence structured data format, and visualization functions of loading of basic scenes, scheduling and displaying of space-time objects, expressing and analyzing of dynamic characteristic information and the like of the patient information visualization are realized.

2) Visualization functions and view design.

The patient information visualization function design comprises basic scenes, object management and scheduling, basic information display, dynamic characteristic expression and other contents, wherein the basic scenes are built and displayed by adopting an osgEarth digital earth framework, and a good three-dimensional display scene can be provided for multi-dimensional characteristic information display and expression of patients and dynamic spatiotemporal process simulation and backtracking. The object management and scheduling mainly realize the loading, management and global scheduling of object information and provide a data basis for the visualization of patient information. Patient basis information presentation patient characteristic information is presented primarily from multiple dimensions. The dynamic characteristic expression mainly realizes the dynamic characteristic information expression of the patient in a spatiotemporal process mode, such as a track moving process, a dynamic attribute updating process, a space form changing process and the like. Finally, backtracking and jumping of the dynamic characteristic space-time process are realized in an interactive mode.

The visual view design content comprises a patient management and interactive view, a basic information display view, a scene display and interactive view, a time management and backtracking view and the like, wherein the patient management and interactive view mainly realizes interactive loading, management and list display of objects, and the main display content comprises a patient time-space domain, a class template, a tree list and the like. The basic information view mainly realizes the display of the object multi-dimensional basic characteristic information, such as the contents of the space-time reference, the space position, the attribute characteristic, the space form and the like of the patient. The scene display and interaction view is a main view of patient information visualization, contents such as spatial form, spatial position, spatial relation, space-time dynamic process and the like of a patient are visually displayed from a geographic view angle by relying on a digital earth frame, and interaction and control with a scene are realized through mouse operation, such as scene stepless zooming and rotation, object selection and query, a right mouse button function menu and the like. The time management and backtracking view mainly realizes the control of the system simulation time and the backtracking of the object space-time process in an interactive mode, such as fast forward and rewind, acceleration and deceleration, start and stop, quick setting and the like of the system simulation time.

3) And visually expressing the flow design.

The patient information visual expression process can be divided into two parts of data loading and visual expression. As shown in fig. 5, the patient information is stored in the spatiotemporal object database, and the data loading process requires the user to log in the system to obtain the corresponding access right and synchronously obtain the object list index information. And on the basis of the object list, a user loads object information as required through interactive operation, and stores the loaded object in a memory object manager, thereby providing support for overall scheduling and dynamic updating of the visual scene object information. In addition, loading of patient information can be achieved through multi-granularity space-time object exchange format file reading, but the process is lack of flexibility and not beneficial to user interaction. The visualized expression can directly call the content in the object manager, and the contents of basic information display, dynamic characteristic expression, scene display and update, full life cycle management, temporal-spatial process backtracking and control and the like of the patient are realized.

4. And performing visual analysis on the displayed content according to actual requirements to obtain visual results of different layers.

The visual analysis in this embodiment includes a macroscopic level visual analysis, a mesoscopic level visual analysis, and a microscopic level visual analysis, and the process of the visual analysis of each level is described in detail below with specific examples.

1) Macroscopic level visual analysis.

And the macroscopic visual analysis adopts an equivalent region method and a partition statistical chart method to establish a spatial-temporal evolution process of the patient and dynamically express the characteristic information of the patient. Taking the statistics of Chinese COVID-19 patients as an example, the data time window is as follows: 2020.01.13-2020.02.25. The main view shows the confirmed diagnosis, cure, death and accumulated confirmed diagnosis conditions of the patients in each province and city in the form of a three-dimensional dynamic thematic map, and the system simulation time step is day and is synchronous with the information acquisition period of the patients. The three-dimensional thematic map constructed by relying on the three-dimensional digital earth frame is different from the traditional two-dimensional map, the stepless zooming and the rotation of the scene can be realized, the mutual shielding and the capping between the objects are avoided, and the characteristic information of the patient is displayed in all directions. The patient-related characteristic information in the basic information view is synchronously updated along with the change of the system simulation time. In addition, the backtracking and analysis of the time-space process of the statistical information of Chinese patients can be realized by interacting time management and backtracking views. Through interactive visual analysis, the number of patients diagnosed in Henan, Hunan, Guangdong, Zhejiang and other places except Hubei is large, and the patients are greatly influenced by the epidemic situation and have high epidemic risk. The spatial and temporal evolution process of the spatial influence of the epidemic situation generally follows the first geographic law, takes Hubei as the center, and the influence range and the influence capacity of the spatial influence process gradually decrease from the center to the periphery and extend to the direction of facilitating traffic travel, such as Guangdong, Zhejiang and Beijing.

2) Visual analysis of mesoscopic level.

The visual analysis of the mesoscopic level adopts the forms of simulating an object track space-time process, a space form space-time process and an attribute space-time process to dynamically express the characteristic information of the patient, wherein the track space-time process describes a space moving process of the patient for a period of time before the patient is diagnosed. The space-time process of the spatial morphology expresses the change process of the icon morphology in the process of patient before diagnosis-suspected-confirmed diagnosis-cured/dead. The attribute spatiotemporal process represents the dynamic change process of the basic characteristic information of the patient, and is automatically updated through an object label or a basic information view. Different spatiotemporal processes are independent, and the types and the number of the spatiotemporal processes which can be added by the same patient are not limited. All the time-space processes can realize backtracking and jumping through interaction with time management and backtracking views, dynamic characteristic information of objects is displayed in the whole process, and a user is helped to improve cognition on epidemic propagation rules.

Taking the example of visualization of COVID-19 patient information in Henan province, the data time window is: 2020.01.17-2020.03.28, and the partial visual effect is shown in figure 6. The object space-time process is mainly displayed in the main view, wherein the trajectory space-time process is assisted by adding a trajectory line for identification, and the historical trajectory of the object can be visually displayed. The hierarchical relation between different objects in the view is controlled through viewpoint height constraint, the track moving process and the space form changing process of the objects are mainly displayed at a high viewpoint, and the attribute changing process of the objects is mainly displayed at a low viewpoint. The initial simulation time step of the system is changed into seconds, and a user can interact with time management and backtracking view to adjust according to actual needs. Through interactive visual analysis, most of patients with COVID-19 in Henan province have the sojourn history in Hubei province, and the ratio is over 60% according to incomplete statistics. Before 23/1/2020, patients in Henan province mainly have foreign input types, and the local infection type is mainly added later. With the increase of the number of infected people and the aggravation of the illness state of the patient, the spatial behavior of the patient gradually moves to a fixed-point hospital, and the number of the hospital receptions and cures gradually increases day by day.

3) Microscopic level visual analysis.

The microscopic level visual analysis is mainly used for adding contents such as patient space traceability analysis, epidemic situation risk assessment and the like on the basis of the mesoscopic level visual analysis, and deep mining and visual analysis are carried out on the patient information by adopting a method for simulating the time-space change process of the patient infection link and analyzing hot spots of the infection-prone area. The infection link is a time-sequential directed graph. The hot spot analysis of the susceptible region mainly shows the hot spot region which is remained in a patient in a period of time before the diagnosis is confirmed in a dynamic thermodynamic diagram mode, and the place with darker color is the susceptible region.

For example, with the visualization of COVID-19 Schulland associated patient information, the data time window is: 2020.05.07-2020.05.23, and the partial visualization effect is shown in figure 7. The system main view mainly shows a dynamic thermodynamic diagram of a patient time-space process and a resident place of the patient before diagnosis, wherein nodes in the time-space process of an infection link are the patient, the node relations are connected through a three-dimensional directed radian, and the node relations can be linked with other time-space processes of an object in a scene to dynamically express characteristic information of the object. The dynamic thermodynamic diagram adopts red, yellow and green three-color gradual change to express the heat of a resident place of a patient, a red area represents a severe susceptible area, yellow represents a moderate susceptible area, and green represents a mild susceptible area. The basic information view is added with an infection relation list on the original basis, and a user can interactively control whether the user is visible or not to assist in analyzing the characteristic information of the patient. Through interactive visual analysis, the time-space process of patient infection can be intuitively reproduced, and a more intuitive reference basis can be provided for epidemic prevention and control. The whole Schulren related epidemic situation is most affected in Jilin city, the number of confirmed persons is also most, and the occupation ratio is up to 56%. Through dynamic thermodynamic interpretation, high-risk areas (susceptible areas) in Jilin cities mainly comprise a four-season garden district, an early city near the river, a new village, a street district and the like, and are the areas of major concern for prevention and control of the whole epidemic situation.

The invention constructs a multi-granularity case space-time object by collecting, processing, analyzing and objectifying the case information, describes and analyzes the case information in the form of the space-time object, and designs and realizes the three-dimensional dynamic visual expression of the case information on the basis. Finally, the case information is visually analyzed and summarized into 3 levels of macroscopical, mesoscopic and microscopic levels, and the rationality of the case information is verified through a Chinese case statistical information visualization scene, a Henan case information visualization scene and a Schuland related case information visualization scene. In general, the method has good portability and good expansion performance, and can help users to quickly find out the potential propagation rules and carry out epidemic risk assessment.