Supporting epidemic intelligence, personalised and public health with advanced computational methods

Supporting epidemic intelligence, personalised and public
health with advanced computational methods
Joao Pita Costa, Daniela Paolotti, Flavio Fuart, Primož Škraba,
Evgenia Belayeva, Inna Novalija
joao.pitacosta@ijs.si
Delavnica EM-Zdravje, 12. Oktober 2015
Uporaba naprednih računskih metod za podporo
osebnega in javnega zdravstvenega varstva ter
nadzora izbruhov nalezljivih bolezni

‣ basic and applied research
‣ 30+ years of history
‣ Torino & New York
‣ ~40 researchers
‣ mainly supported by:
• bank foundations
• EU research grants
• industrial partnerships
‣ focus on:
• complex systems science
• network & data science
• computational social science
• epidemiology & public health
ISI Foundation
www. i si . i t

‣ basic and applied research
‣ 17+ years of history
‣ ~40 researchers
‣ mainly supported by:
• national grants
• EU research grants
• industrial partnerships
‣ focus on:
• data analysis on text,
• web and cross-modal data,
• scalable real-time data analysis,
• visualization of complex data,
• semantic technologies,
• language technologies and
• sensor networks.
Artiﬁcial Intelligence Lab
ai l ab. i j s. si

14 FP6 projects
28 FP7 projects
4 H2020 projects in 2015
14 on-going EU-funded projects
Top 10 commercial partners
Accenture Labs: enterprise search, visualization
Bloomberg: Web analytics, activity prediction
British Telecom: alarm root cause analysis, failure prediction, anomaly detection
Telefonica: real-time monitoring of brands and competition
Google Labs: opinion detection in Google News
Microsoft Research: product expert detection, complex data visualization, text mining consultancy
New York Times: real time web analytics, trend detection
Siemens: “Enterprise 2.0” consultancy
UN FAO: information integration
Wikipedia: controversial topics opinion detection

Quintelligence: large scale analytics -> New York Times, Bloomberg, British Telecom
Cycorp Europe: semantic technologies -> US government, hospitals, …
LiveNetLife: social computing (web presence) -> Accenture, British Telecom
Modro Oko: complex data visualization -> Accenture, UN FAO
Envigence: environmental intelligence -> Several municipalities in Slovenia and Italy
Spinoffs/Startups
Applications
Collaboration: Accenture , Bloomberg, British Telecom, Microsoft Research, New York Times,
Siemens, UN FAO, Google, Yahoo, Oracle, BT, Telefonica, France Telecom, Wikipedia, Stanford
University, University College London, Jožef Stefan International Postgraduate School,
Quintelligence, Cycorp Europe, LifeNetLive, Modro Oko and Envigence.
Inclusion os National companies in EU projects: Amebis, Hermes Softlab, eParticipacija,
TECOS, Kogast Grosuplje, EMO Celje, Avtomobilski grozd, Orodjarski grozd, Quintelligence,
Valji Štore
Patents: J. Leskovec (MSR), M. Grobelnik & B. Fortuna (BT, in process), D. Mladenic & M.
Jermol & M. Grobelnik (Cycorp, in process), M. Jermol & M. Grobelnik & D. Mladenic & B.
Fortuna (Envigence, in process)

INFLUENZANET
40Kparticipants
More than150Ksymptoms
reportseveryyear
‣ 2003: 2 countries
‣ 2015: 10 countries

Online Influenza Surveillance
Influenzanet providing weekly information on Influenza related topics, the
real-time information on the evolution of the disease at the local level, and
some statistics on the available data.
Influenzanet.eu

Real-time Data Analytics
Fluoutlook: activity data and forecasts are updated weekly, based on the
reports of the official influenza surveillance systems in each country. The
data sources also include Influenzanet and Twitter.
fluoutlook.org

Computational Epi. Models
GLEAM produces realistic simulations of the global spread of infectious
diseases. It integrates three layers: real-world data on the global
population; real-world data on the mobility of this population; an individual
based stochastic mathematical model of the infection dynamics.
gleamviz.org

Health Media Monitoring
3XL News provides instantaneous global news monitoring and analysis
across several languages on-the-go, showing how semantic technologies
are used in a real-world scenario.
ailab.ijs.si/tools/3xl-news/
ailab.ijs.si/tools

Social Media Nowcasting
The portal twitterobservatory.net to extract information from Social Media
(Twitter). In this case the query was influenza.
twitterobservatory.net
ailab.ijs.si/tools

Cross-lingual Technology
While XLing is a Wikipedia based cross-lingual similarity computation
across 100 languages, NewsFeed is a real-time aggregated stream of
semantically enriched news articles, and EventRegistry is a system for real-
time collection and analysis of news published globally.
ailab.ijs.si/tools
newsfeed.ijs.sixling.ijs.sieventregistry.org

Early Disease Detection
eventregistry.org
23 March 2014
11,312 people reported infected
6 countries: Liberia, Guinea, Sierra
Leone, Nigeria, USA and Mali.

The portal searchpoint.ijs.si exhibiting the clustered keywords of a query, after
searching for the keyword influenza. The portal scienceatlas.si exhibiting the
graph of academic collaboration related with the query influenza.
Visualization of Complex Data
searchpoint.ijs.si scienceatlas.si
ailab.ijs.si/tools

Topological Data Analysis applies the qualitative methods of topology to
problems of machine learning, data mining and computer vision.
Persistent homology identifies a global structure by inferring high-
dimensional structure from low-dimensional representations and studying
properties of a (often) continuous space by the analysis of a discrete
sample of it, assembling discrete points into global structure. It has
important applications in Medicine including Cancer research.
InfluenzaLab.org
Topological Data Analysis
Tumor Detection

Influenzanet SIovenia
ailab.ijs.si
Hospital admissions
Clinical reports
Online volunteers
Extension to other acute
respiratory infections [ARI]

A/C Respiratory Diseases
ailab.ijs.si
Number of self-reported illness from a survey in Portugal 2011-2013
ASTHMA + COPD + ALLERGIC RHINITIS
prevalence of asthma in adults and
children is between 1 and 18% [WHO]
61+ million Europeans affected
by allergic rhinitis
susceptibility of immigrants and
refugees to pulmonary diseases
The nr. of people aged 60+ (COPD target gr.)
is now increasing by about 2 million per year
The Portuguese Influenzanet

ailab.ijs.si
Big Health Data
cross-modal big data fusion
microbione profiling of sputum samples
swabbing of inline volunteers
micro-environmental exposure sensor signals
small RNA profiling peripheral blood samples
Influenza net real time data
feature extraction
semantic representation, correlation, and alignment
indexing and retrieval
nowcasting of social media
topological data analysis
Laplacian eigenmaps
canonical correlation analysis
Bayesian networks for large scale prediction
Monte Carlo simulation on gene expression data
Auto regression and Gaussian models
correlation preserving mappings
generalised multi-view analysis
multidimensional scaling
report on real-time data
reliance on self-reported data

Health Platforms & Gadgets
Apple Health Microsoft HealthVault Google HealthIBM Watson Health
FitBit Runtastic Samsung Digital Health Telefonica

Internet of Health Things
AILAB

Wrap-up time!!
EPIDEMIC
INTELLIGENCE

Computational Topology Data Visualization Gripenet
Social Media Mining Medical Data Analysis Persistent Homology
Python Influenza Like Diseases AI Lab Big Data Analytics
TDA Epidemic Intelligence Digital Disease Detection
ISI Foundation Artificial Intelligence Data visualization
AI Lab Respiratory Diseases Computational Algebra
HVALA

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
BE
NL
PT
IT
UK
FR
SE
ES
Inﬂuenzanet
Europe-wideparticipatorysurveillance
inﬂuenzanet.eu
D.Paolotti et al.,
Clinical MicrobiologyandInfection,20(1),17(2014)

GLEaMviz - Global Epidemic and Mobility Model
gl eamvi z. or g

SensorLab
Cross-departmental lab for
wireless sensor networks
AILab
Department of
Communication
Systems
Centre for Knowledge
Transfer in Information
Technologies
Goal: combine technologies for sensor data acquisition, communication between sensor devices,
statistical real-time data analysis, semantic technologies, and to enable a wide range of research and
development in different application areas (energy, ecology, health, transport, security, logistics).

Supporting epidemic intelligence, personalised and public health with advanced computational methods

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (18)

Similar to Supporting epidemic intelligence, personalised and public health with advanced computational methods

Similar to Supporting epidemic intelligence, personalised and public health with advanced computational methods (20)

Recently uploaded

Recently uploaded (20)

Supporting epidemic intelligence, personalised and public health with advanced computational methods

Editor's Notes