We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

A framework for assessing location-based personalized exposure risk of infectious disease transmission

00:00
subtitles
off
playback rate
1
subtitles
off
English (auto-generated)
playback rate
0.25
0.5
0.75
1
1.25
1.5
1.75
2
quality
576p
11
 Cite
 Share
 Download Purchase
No logo availableFOSS4G
Logo of Open Source Geospatial Foundation (OSGeo)Open Source Geospatial Foundation (OSGeo)
 Ching-Shun, Hsu

Formal Metadata

Title
A framework for assessing location-based personalized exposure risk of infectious disease transmission
Title of Series
Number of Parts
183
Author
License
CC Attribution - NonCommercial - ShareAlike 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this
Identifiers
Publisher
Release Date
Language
Producer
Production Year2015
Production PlaceSeoul, South Korea

Content Metadata

Subject Area
Genre
Abstract
Human mobility is an important risk factor affecting disease transmission. Therefore, understanding detailed spatial behaviors and interactions among individuals is a fundamental issue. Past studies using high-resolution human contacts data from smart phones with GPS logs have captured spatial-temporal heterogeneity and daily contact patterns among individuals. However, measuring personalized exposed risk of infectious disease transmission is still under development. The purpose of the study is to establish a location-based framework for assessing personalized exposed risk of infectious disease transmission. The framework consists of three components: the first is client-side smart phone-based risk assessment module. We developed Android application for collecting real-time location data and displaying the personalized exposed risk score. The second component is the server-side epidemic simulation model. The simulation model calculated the personalized exposed risk score based on real-time GPS logs and individual mobility data from the client-side Android application. The last component is the disease alarm device for triggering the service-side epidemic simulation model. We installed infrared sensors in people-gathering areas as the alarm device to monitor human body temperature for detecting fever syndrome. We used NTU main campus as a pilot study to demonstrate the feasibility of the framework. We analyzed the records of students’ taking course and modeled the spatial interaction relationships among classroom buildings due to students’ mobility around the campus. Someone who got a fever is detected by the sensor and the server-side epidemic simulation is triggered. Each student who installed the client-side risk assessment module in his/her smart phone receives the real-time personalized exposed risk score when an epidemic outbreak on the NTU campus. The study proposed a location-based framework for measuring real-time personalized exposed risk. Each student at the campus could understand the spatial diffusion of disease transmission and make better spatial decisions based on personalized exposed risk scores.
00:00
Software frameworkData transmissionBenz planeEmailStudent's t-testIntegral domainInflection point19 (number)Gamma functionObservational studyImage resolutionIntegrated development environmentVideo trackingPattern languagePoint cloudInformationComputer simulationBlogExplosionPlastikkarteBuildingElectronic data interchangeAreaMIDIConvex hullSocial class2 (number)INTEGRALSpacetimePerspective (visual)Statistical hypothesis testingFunctional (mathematics)Source codeResultantStudent's t-testBuildingRow (database)RoutingElement (mathematics)PlotterImage resolutionSoftware frameworkMobile appMultiplication signData transmissionObservational studyMobile WebGoodness of fitPresentation of a groupSocial classCoroutineFocus (optics)Variable (mathematics)Service (economics)Interactive televisionSoftware developerCartesian coordinate systemSimulationIntegrated development environmentComputer simulationPattern languageLoginDatabaseQuery languageState of matterPerturbation theoryDependent and independent variablesNoise (electronics)Moment (mathematics)Sound effectCausalityUniverse (mathematics)Greedy algorithmPhysical lawConnectivity (graph theory)Rule of inferenceSystem dynamicsTape driveControl flowProgram slicingAnalogyInformationGradientPressureNumberAreaComputer animation
06:24
Computer networkMatrix (mathematics)Computer multitaskingSocial classExecution unitBuildingStochasticMeta elementComputer simulationWeb pageForceControl flowStrategy gameCurveExplosionInformationPlastikkarteNewton's law of universal gravitationPoint cloudQuery languageGamma functionGame theoryImage registrationService (economics)BlogDatabaseElectronic data interchangeTime zoneStudent's t-testDisintegrationSummierbarkeitSpecial unitary groupDecision theoryOnline helpIntegrated development environmentPattern languageHill differential equationSoftware frameworkObservational studyData transmissionEmailIntegral domainTwin primeRule of inferenceBuildingPrice indexInsertion lossService (economics)DatabaseInformationSpacetimeTraffic reportingDirected graphIntegrated development environmentType theoryReal numberHypermediaStrategy gameWebsiteGame controllerCasting (performing arts)Pattern languageComputer simulationMobile appQuery languageData transmissionSimulationSoftware frameworkStudent's t-testDecision theoryAnalogySystem dynamicsState of matterMultiplication signOperator (mathematics)Network topologyArithmetic meanStatistical hypothesis testingForcing (mathematics)Presentation of a groupFrequencyCurveTheory of relativityProgram slicingTopological vector spaceInteractive televisionPerturbation theoryFunctional (mathematics)SoftwareSystem callPower (physics)SmartphoneControl flowRow (database)Android (robot)LoginMatrix (mathematics)Mobile WebImage registrationSubject indexingSlide ruleAuditory maskingData miningStochastic processRoutingOverlay-NetzServer (computing)Message passingInternet service providerComputer animation
12:44
19 (number)Student's t-testExecution unitDisintegrationForceLemma (mathematics)MiniDiscNumberStatistical hypothesis testingVertex (graph theory)CASE <Informatik>Image resolutionPoint (geometry)Rule of inferenceStudent's t-testView (database)CalculusCAN busBuildingAreaVideo gamePopulation densityMaizeClosed setSubject indexingError messageRootComputer animation
16:39
DisintegrationForceoutputComputer animation
Transcript: English(auto-generated)
00:04
Good afternoon, ladies and gentlemen. My name is Xu Jingshen, come from National Taiwan University. The topic of my presentation today is a framework for assessing location-based personalized exposed risk of infectious disease transmission.
00:21
When we try to understand the relationship between our health outcome and environmental exposure, more and more studies focus on personal exposure assessment. Because where and when people spend their time differ from individual to individual,
00:41
it is important to assess exposed risk based on their behavior patterns instead of where they live. Meanwhile, with the development of GPS, some wearable sensors or personal monitoring device, we can collect personal,
01:01
we can collect personal mobility data and the environmental data more accurately. And integrate with GIS to estimate personal exposure risk. Some studies about these issues are talking about,
01:33
sorry, some study about these issues
02:04
are talking about air pollution exposure, or pesticide exposure, and so on. And in epidemiology, most of the studies with analyzing human mobility data try to understand collective behaviors
02:20
and the special diffusions of infectious disease. For example, how does a disease spread between cities? Or what is the impact of variable mobility routines on infectious disease dynamic? They seldom focus on the perspective on personal exposed risk of infectious disease.
02:43
However, there are some distinctive features between infectious disease and environmental pollution. For example, your risk will be affected by the people surrounded with you. The more people, the more sick people surrounded with you,
03:02
the more risk you are. That is the source of infection could go faster than environmental pollution. On the other hand, the impact of the disease is more serious and quickly. To sum up my introduction,
03:22
with GPS or positioning function in a wearable device, we can collect high resolution individual space-time data. Based on this data, we are about to understand the personal environmental pollution exposure. But in epidemiology, the perspective
03:41
on personal exposure assessment of infectious disease is still unclear. So the purpose of this study, the purpose of this study is to establish a framework for assessing personalized exposed risk assessment of, to assessing personal exposed risk
04:04
of infectious disease transmission. The framework consists of two components. The first is client-side smartphone application. We developed Android application for collecting personal GPS logs data
04:21
and providing users with personal exposed risk assessment service. The second is server-side epidemic simulation model. The model simulates the spatial diffusion of disease and integrate with personal mobility data
04:40
to calculate the personal risk score. In this model, we analyze students' course-taking records to understand how the students work on the test results. Sorry, we use school campus as a plot study
05:00
to demonstrate the facility of this framework. We analyze students' course-taking records to understand how the students work on the campus and to model the spatial interactions between classroom buildings. And with this relationship, we can model
05:22
the spatial diffusion of the disease on the campus. That is my study area, National Taiwan University main campus. There are 75 classroom buildings. Next is students' element database.
05:42
With query function, we can know the students' course-taking records from individual to individual. For example, which course he takes and when and where to go to the class. For example, the students will go to gymnasium
06:03
on Monday morning and then go to building A and from building A to building B and so on. So with this information, we can analyze his route, routes on the campus. And we divide these routes into two time slides,
06:22
morning and afternoon. Then we aggregate all of the students' routes on the campus to establish the building's network. It is a origin destination matrix for each time slide. So we can understand the spatial interactions
06:43
between classroom buildings per half-day. And we integrate the relationship with meta-operation approach and the IR stochastic model to simulate the spatial diffusion of disease on the campus.
07:02
With the model, we can know the dynamic infection patterns by building's peer half-day. And based on the definition in epidemiology, the force of infection, we calculate this index as a risk score for each building.
07:21
So far, we can understand the dynamic of disease diffusion and can answer some questions. For example, with plotting the epidemic curve of this outbreak, we can realize how long do we have to plan the disease control strategy.
07:41
By visualizing the risk score of each building, we can know which building is the most serious one at a specific day. And to point out the other buildings that have cross-relationship with. Cross-relationship means there are a lot of students
08:03
come and go between two buildings. If one building's outbreak and the disease would spread to the other one quickly. And based on this information, we can plan the strategy more precisely.
08:21
Next, we are going to look at the function of the smartphone application. This app is developed in Android environment, which consists of three functions, including registering an account,
08:42
collecting personal GPS logs data, and the clearing personal risk of infectious disease transmission. First, registration is to create an account in database and provide users with risk assessment and alert information service.
09:03
Once there is a CAS report on the campus, the app will receive a message from the server to tell the users there is a disease outbreak. Secondly, we collect personal GPS logs data
09:20
with positioning functions on the smartphone and upload this data to the database. Meanwhile, we developed a simple GPS log data mining method to guess which buildings the user has visited. Otherwise, if he registers with his student's ID,
09:43
we can also know his behavior patterns on the campus by visualizing his cross-checking records aforementioned. And with simulating the spread of disease, we can understand the spatial patterns of each,
10:02
of exposed risk, and overlay these patterns with personal rules on the campus. Based on this idea, we developed an indicator to measure how dangerous the user is as a personal exposed risk score.
10:24
Finally, at the query user interface, we decide two types of query. The first one is campus query for environment risk. When a disease outbreak, as a student, he may want to know how dangerous is the buildings
10:43
he is going to, and with the query, users can understand the spatial patterns of the risk on the campus. On the other side, a personal query provides the risk score information to each users.
11:01
The risk score is based on their own mobility patterns. They are about to know the risk they have encountered or are going to face. We suppose this information can help users
11:21
make better spatial decisions, such as avoid visiting high risk place to reduce the probability of getting sick, or if he doesn't feel well, and he found that he has exposed to high risk,
11:43
and it is better for him to see a doctor or wear a sanitary mask.
12:08
In conclusion, we propose a framework for personal exposed risk of infectious disease transmission, which consists of smartphone apps and epidemic simulation model.
12:21
With information about the environmental risk or personal risk, we may make better decisions when a disease outbreak. Thank you.
12:40
Any questions? So how would you identify patient number one? So how do you know when there's an infectious disease
13:02
in the area that you would start deploying this scenario? On the campus, there is health centers where students can think they would go to the centers firstly
13:24
to see the doctor in the school. Once the students is once the students get disease is
13:45
infectious disease, and we suppose that the center would report this case to the school
14:03
or CDC and so on. Any else question?
14:20
Any other question? Last is me. I have a question about exposure risk score today. Yes, yes. Risk score. Yeah, yeah.
14:41
So is it how to make a value to the risk score? How to? Make a value. Is it the numbers? 10 or 11 or 13, 14, 19, the risk score?
15:02
Is this? We think this score is relative index. So he he can know which few things is more dangerous one. And we compare the test he was going to.
15:21
He can choose the most less serious one. Yes, go on. Is it the red one is a high-risk point? Yeah, yeah. No, no, no. Red point.
15:45
This one, this one. The error is the personal roots on the campus. And the note is buildings.
16:01
So this is first these buildings on the Monday morning. And we also know each building's risk score. So we can know which buildings he has visited. And we calculate
16:21
the risk of the risk of the buildings to calculate his personal risk score. Okay. I understood. Thank you. Thank you. Thank you very much.
16:43
Thank you. Thank you very much.