Download Network Based Approaches - Global Digital Health Network

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Network Based Approaches to
Studying Ebola and Emerging
Infectious Diseases in Real-time Using
Smartphones
Solomon Abiola, MS
CHET/DSAIL
Outline
Rationale and background
mHealth and the application of Node to Ebola
Network based approaches to combat disease
2
Twitter: eWizard2_0
www.oneabitech.com
Mobile health and big data among the current most
hyped technologies in 2014
3
Source: Gartner
Twitter: eWizard2_0
www.oneabitech.com
The rapid growth of mobile technology allows it to
serve as the perfect health platform
Why mobile health?
There are over 334 million smartphones projected to be present in Africa in 2017
~ 30% of the continent’s population
How can we use mobile health?
Remote assessments, diagnostics, increase access to care, disease modeling,
etc.
4
Source: TechCrunch
Twitter: eWizard2_0
www.oneabitech.com
Objective real time geospatial modelling via
smartphone sensors improves contact tracing
5
Twitter: eWizard2_0
www.oneabitech.com
The Princeton University pilot allowed us to
geospatially identify infectious disease sinks
Main Application Page
Source: Pilot Study
Twitter: eWizard2_0
www.oneabitech.com
Areas highly susceptible to interaction among students with the
application appear darker. (Aerial Map of Princeton University)
Node enables health workers and practioners
to see incidents in real time
Two user interactions sampled from March 17th, 2015
Twitter: eWizard2_0
www.oneabitech.com
The Node study is designed to recruit a small subpopulation in Lagos, Nigeria
Through our preparatory activities we’re expecting at least 100 participants (>40
University of Lagos + > 25 Nigerian Institute of Medical Research + >40 A/B testers)
Enroll up to 100 individuals within Lagos state (city population over 21 million)
Participants will be paid 50 USD equivalent in phone credits per month of
completion during the study
8
Twitter: eWizard2_0
www.oneabitech.com
Knowledge Discovery Process has lead to
interesting insights in mHealth for disease
mHealth App
Node
~70 users
Parse
(Facebook)
Google
CartoDB
PHASE 1
Twitter: eWizard2_0
www.oneabitech.com
MongoDB
~30GB
AWS
PHASE 2
PHASE 3
Analytics
~2mb
PHASE 4
9
The Node View platform allows for real time
health action based on ubiquitous data
Locate
Identify
View
Enact
Lagos, Nigeria Beta Test
Twitter: eWizard2_0
www.oneabitech.com
10
Three distinct networks emerged over the
course of the study, correlated with study pop.
11
Twitter: eWizard2_0
www.oneabitech.com
Network evolved over three month time
period, with some nodes leaving and coming
AUG
SEP
OCT
12
Twitter: eWizard2_0
www.oneabitech.com
Worse time to contract diseases starts early in
the week – e.g. Today!
Calendar Heatmap of contact events for entire study duration.
13
Twitter: eWizard2_0
www.oneabitech.com
So if we perform a targeted removal of highly
infectious individuals what happens?
AUG
Graph Density (.127  .098)
Edges (198  130)
Twitter: eWizard2_0
www.oneabitech.com
SEP
OCT
14
Public Health Official – “And so what?”
Actionable data. Finding patient ZERO.
When?
Calendar Heatmap of contact events for select user.
Who? Subject “_User$ri48ujTmJK”
How? Where?
15
Twitter: eWizard2_0
www.oneabitech.com
Where? Using geospatial information patient
ZERO can be located, and path protected.
16
Twitter: eWizard2_0
www.oneabitech.com
Where? Using geospatial information patient
ZERO can be located, and path protected.
_User$ri48ujTmJK
Activity
in_vehicle
on_bike
on_foot
still
tilting
64
unknown
34
4
49
580
96
17
Twitter: eWizard2_0
www.oneabitech.com
Beyond Ebola, future emerging infectious
diseases
Continuous temperature tracking
Household influenza tracking
Disease forecasting
NSF #1516340
[1] Abiola, S. O. Node: A Real Time Smartphone Big Data Application for
Health Care Epidemiology. (2013). at
http://dataspace.princeton.edu/jspui/handle/88435/dsp01sx61dm401
[2] Abiola, S. O., Portman, Eric., Kautz, Henry., Dorsey, E.R., Node View: A
mHealth Real-time Infectious Disease Interface. Ubicomp/ISWC’15 Adjunct
Proceedings. (2015).
Twitter: eWizard2_0
www.oneabitech.com
18
Emerging infectious diseases platform –
example case of Zika
19
Mobile Development and Analytics Courtesy of AbiTech, Inc. www.oneabitech.com
Twitter: eWizard2_0
www.oneabitech.com
20