1 Digital Catapult The impact of Big Data in a Connected Digital Economy Future of Healthcare Mark Wall Big Data & Analytics Leader March 12 2014 Catapult is a Technology Strategy Board programme
Agenda 2 Awareness Building Recap on what is Big Data Second Wave of Innovation EXAMPLES FROM THE FIELD Data Driven Research Innovation - Now Tackling Dengue Fever APAC Biogrid Australia Second Wave of Health & Disease Intervention Innovation Summary & Call to Action
But first Social Media saves girls life 3 Girl, 2, saved by a photo on Facebook after family friend spots eye cancer in photo taken by mother
Social Media saves girls life 4 Girl, 2, saved by Facebook after family friend spots eye cancer in photo taken by mother
Recap: Three Driving Forces 5 Social + cloud + mobile = Big Data
Hype has Created Many Misconceptions about Big Data 6..not new..not about Tech 1989 Harper Magazine publication Doug Laney, Gartner 2001-4 V s in big data. Was all about Data Integration New Business Opportunities Transformational Business and Processes..IS about Data Shift in relation of Data to science and business
Three Emerging Drivers Important for Healthcare 7 sensor + nano + bio = Big Data Sensors are really big data Unbounded, Promiscuous, Indiscriminate Internet of Things is bigger Everything is connected; everything communicates; everything is a sensor Inanimate becomes Sentient
Quantifying Big Data in Healthcare 8 Sources: medical records, citizen data, genetic data, radiology images, nano sensor data Current trajectory from 2012 is scaling 500 to 25,000 petabytes by 2020 New Opportunities to improve outcomes, reduce cost of delivery
Impact 10 Datafication of human intention and processes Data analysis replaces data acquisition as new bottleneck Mobile Sensor platforms collecting data Traditionally: Query the world Data acquisition activities coupled to a specific hypothesis BigData: Download the world Data acquired en masse in support of many hypotheses
Social + Mobile + Cloud has 11
What this means for Future of Health? 12 Disease Prevention Industry Anticipating and predicting Health better Within the Individual and in the Community Connectivity: Sensors + Nano + Bio Real time streams of Information directly from Sensors, nano technology Within the human body Broadcast Human Data Streams Environmental Factors, what s trending Internet of Things #IoT
THIRD PARTY COMPANY LOGO A signification portion of clinical data is not yet digitized even in modern economies. There is a substantial opportunity to create value if these pools of data can be digitized, combined and used effectively. Mckinsey Global Institute Big data: The next frontier for innovation, competition, and productivity, May 2011
14 Examples of Big Data Innovation Now
Tackling Dengue Fever in Singapore Social Media + Mobile + Cloud 15 Geomedical Systems: Intervention and Control (Richard Thomas, University of Manchester) Datafication, data enrichment of epidemiological model Complex interaction of environmental factors with vector life-cycle Example of Collaborative Social Intelligence
Tackling Dengue Fever Solution - Deploying next best anticipated action in the cloud 16 Lambda Architecture Continuous Streaming Analytics Deep Analytics on all Data combined Streams of Events Disease Intervention Recommendations Event Processing Engine Model Parameter Optimizations Events NoSQL Big Data Analytics Engine Hadoop DW Graph / linked Data Model Latency m/s Rules seconds < 5minutes
Data Mashups & Trusted Personal Data: BioGrid Australia Leading the Way in Collaboration 17 Not for profit, Operates secure research infrastructure since 2003 Provides access to real-time clinical, imaging and biospecimen data Unlock jurisdictions, institutions and disease specialisms Web-based platform provides ethical access Protects privacy & intellectual property Winner of a 2013 Computerworld (global) Laureate in Health Category
Unlocking and Accelerating Medical Research How BioGrid Australia works 18 Unique Subject Identifier Linkage Key Federator Enables patient data to become anonymous but identified as an individual for the purposes of disease tracking at individual level Automated process enabled via a secure VPN layer Queries are routed through secure layer and brokered across datasets via the linkage key federator (LKF)
19 Next Wave of Innovation Collaborate > Unlock > Accelerate
Next Wave of Innovation: Sensor + Nano + Bio 20 Personalised medicine Tailored genetic mapping to drugs & pathways Ingestible sensor platforms Nanobot delivery vehicles into target pathways and tissues Higher Relevance (Time and context) of Information Lower latency in information delivery Ability to handle higher variety and volume of data
Graph Databases Enabling Personalised Medicine 21 Example of NoSQL Technology Coming to the fore now Nodes and Edges Not Tables and Columns Very fast context driven queries against bigdata Video content
Latency is a killer Support for Streaming analytics, not just Real time 22 Latency breeds contempt Aggressive demanding lines of inquiry are good. We must support them! An example - Data always wins over tools: Sophisticated tools without data are useless Mediocre tools with access to the data are frustrating to use Analysts ALWAYS opt for frustration over futility Follow Up Lambda Architecture for more An Architecture blueprint for bigdata Continuous Streaming Analytics
THIRD PARTY COMPANY LOGO The cost-benefit ratio for traditional DW s is starting to reach the point of Inversion. Understanding the Logical Data Warehouse: The Emerging Practice Mark Beyer (June 2012), Gartner Source: http://www.gartner.com/id=2057915
Example: Big Data & The Cancer Genome Atlas National Cancer Institute Funded Frederick National Laboratory (USA) Cross Reference TCGA gene expression data from simulated 60 million patients with mirna expression data for a simulated 900 million patients Result: Understanding additional layers of the pathways these genes operate in and the drugs that target them. Enabled by lower total cost of processing and parallel processing in Hadoop 24
National Cancer Institute Identifying Relationship between Gene and Cancer Interaction 25 Cross-referenced the relationships between 17000 genes and five major cancer types across 20 million medical publication abstracts Cross-referenced genes from 60 Million patients and mirna for a simulated 900 Million population. Understanding additional layers of the pathways these genes operate in and the drugs that target them is expected to help researchers in their work
Summary 26 Big Data: Opening New Opportunities to tackle and intervene disease faster Social + Mobile + Cloud = Big Data Nano + Bio + Sensor even bigger data Creating new innovation opportunities for health + much more!
Call to Action 27 Collaborate > Unlock > Accelerate At the beginning of a 5 year program Working with Innovators that are facing market challenges to unlock obstacles Create architecture blueprints E.g. Open Health Data Platform
Catapults What 2013 7 Catapults up and running A network of technology and innovation centres Long-term investment Foundation funding by the Technology Strategy Board Why 2013 2 New Catapults in 2015/16 7 Catapults up and running Increase the UK s ability to create new products and services To create sustained economic growth How 1bn Private & public sector investment By building collaborations Having a delivery focus and hands on
Contacts 29 THANK YOU Mark specialises in Strategy & Analytics within the Communications sector, where he is working with IT and business leaders on IT strategy, target operating models, data monetisation and cloud based Analytics as a Service. He has a special interest in driving economic development and raising standards for society using data driven analytics.