Internet of Things Future Vision and Nursing
InvolvementThomas R. Clancy, PhD, MBA, RN, FAANClinical Professor and Associate Dean
Faculty Practice, Partnerships, Professional DevelopmentSchool of Nursing
The University of Minnesota
Objectives
• Describe the relationship between growth in information, complexity and the challenges of big data.
• Define machine learning and discuss the emergence of augmented intelligence.
• Describe the information value loop and how the Intranet of Things enables it.
• Provide current and future examples of how big data, The Intranet of Things and artificial intelligence will change the role of nurses.
Information
Information Theory
• Information is the ordered arrangement of matter that represents something.
• Meaning is not embodied in it. Ordered matter or packets of
information in solids
Humans have the Unique Capacity to:
• To create, recombine, store and process complex ordered states or information in our minds.
• Disembody information into useful artifacts such as books, buildings, roads and other “things” that improve our survival and quality of life.
Ordered matter or packets of information in solids
Information: Bits & Bytes
• Information can also be represented in binary symbols or as bits.
• The ordered patterns of binary symbols represent information.
Ordered matter or packets of information in solids
How Information Grows
Information begets more information:
• Evolution & human cognitive capacity.
• Emergence of networks to share individual packets of information where gaps occur.
• Information technology enables machine aided intelligence
• Super IntelligenceUnique Information is growing
66% per yearVarian, H. and P. Lyman (2003). How Information Grows @
http://kk.org/thetechnium/the-speed-of-in/
Long Term World Growth in Gross Domestic Product/Per Capita
Moores Law & Technology
The Perfect Storm: Mobile Technology, Sensors,
The Intranet & Artificial Intelligence
2020 X Artificial Intelligence
2010 X X X Intranet
2000 X X X X X Mobile Tech
1990 X X X X X
1980 X X X
1970 X X X
Year 1960 X X X X Sensors
1950 X X X
1940 X X X
1930 X
1920 X
1910 X X
1900 X X X X
Mobile
Technology
Mobile
Phone Call
Pager Cordless
Phone
Hand held Mobile
Phone
Commercial Mobile
Phone
Apple PDA Bluetooth Camera
Phone
Blackberry
Smart Phone
Apple
Smart
Phone
Android
Smart
Phone
Sensor
Technology
Temperat
ure Wind Humidity Biosensor Radiation Acoustic Velocity Force Light Motion Position Chemical
Internet ARPANETFTP/TCP/IP/Et
hernet Internet
WWW/Java/Net
scape
Amazon/Google
/eBay
iTunes/Faceb
ook/Napster
Skype/YouT
ube Twitter Instagram Snapchat
Artificial
Intelligence Birth of AI
Rules Based
Systems AI Winter Expert Systems Neural Networks
Evolutionary
Algorithms
Baysian
Networks
Deep
Learning
https://en.wikipedia.org/wiki/Moore's_law#/media/File:Transistor_Count_and_Moore%27s_Law_-_2011.svg
Sources of Data in Healthcare • Electronic Health Record
• Health Insurance Claims
• Sensor Data (2.9 billion)
• Geo-spatial Data (GPS mapping)
• Intranet of Things (IoT)
• Social Media (1.8 billion subscribers – top 5)
• Patient Reported Outcomes (quantified self movement)
• Human Genome (6 billion/pair)
• Financial Systems (credit cards, bank accounts)
• Environmental and Weather Data
Information as it Grows Becomes Complex• Information and
complexity are closely associated.
• As we move along the continuum from randomness to complete order, it is not only the amount of matter present, but the specific arrangement of it that increases information and its complexity.
Randomness OrderComplexity
Low Information
Low Information
HighInformation
Gas Crystal
Information Complexity: Nine Factor Binary Matrix
One Person = 1,024 Combinations
ID Age Ht. Wt. Pul. Inc. Edu. Occ. Lab Xray
Age N Y N Y N Y N Y
Ht. Y Y Y Y N Y N Y
Wt. Y Y N N Y Y Y Y
Pul. N N Y N Y N Y N
Inc. Y N Y N N N Y N
Edu N N N N Y N N Y
Occ Y Y Y Y Y N Y N
Lab N N N Y Y Y Y N
Xray N N Y Y N N Y N
Finding Hidden Patterns
•What is the probability of solving a Rubik’s cube by randomly turning the cubes? 1 in 43,252,003,274,489,856,000
(4.3 x 1019)
324 sides
Machine Learning
• The science and technology of systems that learn from data.
• Used to solve complex problems and describe the structure of the data generating processes.
Data Methods: Machine Learning
Artificial Intelligence:
Machine Learning
• Decision Trees
• Neural Networks
• Bayesian Methods
• Evolutionary Computation
• The goal is to predict rather than explain.
• There are many predictor variables (25+).
• There are many complex variable interactions.
• The predictors have non-linear relationships to the target variable.
DemographicsDiagnosis codesProcedural codesProviderUnitBSNMedicationsNursing HoursLength of Stay
Diagnosis Diagnosis + Procedure CodeDiagnosis + Procedure Code + AgeDiagnosis + Procedure Code + Age + LOSDiagnosis + Procedure Code + Age + LOS + ICUDiagnosis +Procedure Code + Age + LOS + BSN
1. Select concepts
2. Remove irrelevantfactors
3. Search for positiveconjunctive conjectures
Diagnosis + Procedure CodeDiagnosis + Procedure Code + AgeDiagnosis + Procedure Code + Age + LOSDiagnosis + Procedure Code + Age + LOS + ICU
4. Remove negativeconjunctive conjectures
5. Build Algorithmicrules around positive
conjunctive conjectures
• Data Scientist• Informatician• Statistician• Domain Experts• (nurse scientist)
Decision Tree: Predictive Model
CAUTI
Garter Hype Curve:Machine Learning
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
AI Algorithms -> Augmented IntelligenceEmbedded in Devices
Park, J. (2016). Developing a Predictive Model for Hospital-Acquired Catheter-Associated Urinary Tract Infections Using Electronic Health Records and Nurse Staffing Data. Dissertation. University of Minnesota
The Information Value Loop
• Create• Sensors (generate data)
• Communicate• Network (transmit data)
• Aggregate• Standards (gather data)
• Analyze• Augmented intelligence
(patterns and signals)
• Act• Augment (change)
behavior
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
Create
Communicate
Aggregate
Analyze
Act
The Intranet of Things
“The integration of people, processes and technology with connectable devices and sensors to enable remote monitoring, status, manipulation and evaluation of trends of such devices.”
https://www.youtube.com/watch?v=sGQeWRpmglU
Peter Lewis. First use of the term given at a presentation toU.S. Federal Communications Commission (FCC) in 1985.
The Intranet of Things
Smart PhonesRobotics• Surgical• Care-assist• Gero-tech
Desktop Computers• EHR
Implantable Devices• Diabetes• Cardiac
Wearable Devices• Exercise• Sleep• Stress
Drones
MobileTablets
In-homeSensors• Motion• Alarms• Medication
Smart clothing
Medical Devices• ECG• Vital signs• Ultrasound
Telehealth• Virtual
visit
Nanotechnology
Lab/Radiology
Intranet of Things
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
Create: Sensors
Types: • Passive vs active
Growth Factors• Decreasing cost ($2-$.40)
• Improved computation (x2/3yr)
• Decreasing size (3 – 10/phone)
Challenges• Power consumption (battery)
• Security (ongoing issue)
• Interoperability (communication
std’s, proprietary)
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
Communicate: Networks
Types:• Wired vs Wireless
Growth Factors• Data rates (2Kbs analog, 1Gbps digital)
• Intranet transit prices ($120 - $.63 Mbps)
• Power efficiency (Bluetooth Low Energy 50%)
• IPv6 adoption (address conversion IPv4)
Challenges• High bandwidth network
penetration (conversion to 4G limited)
• Security (limited with IP connections)
• Power (increasing power demand)
• Interconnections (http/interoperability)
PAN (personal area network)Bluetooth, ZigBee,
Near Field Communication, Wi-FiLAN (Local area network)
Wi-Fi, WiMAXWAN (Wide area network)
WiMAX, weightless, cellular technologiessuch as 2G, 3G, 4G (LTE)
USB orEthernet
Wired
Wireless
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
Aggregate: Standards
Types of Standards:• Technology and regulatory
Growth Drivers• Network & communication
standards are emerging
• Converging standards
• Vendors & IEEE/ETSI
• Qualcom/Cisco
• Data aggregation standards
• Unstructured data (ETL)
• Security & PHI (HIPPA)
• Data mart standards
• Technical Skills & big datahttps://www.youtube.com/watch?v=qkW66bOlkBU
Network Protocols• Machine identification• Multiple protocols
Communication Protocols• Common language• HTTP & IoT
Data Aggregation• Streaming data• Relational databases• SQL vs NoSQL• Distributed Databases
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M.
Inside the Intranet of Things. A primer on the technologies
building the IoT. Deloitte University Press.
Challenges
A Connected Planet
https://www.semiwiki.com/forum/content/5559-quick-history-internet-things.html
Analyze: Augmented Intelligence
Types• Descriptive (describe)• Predictive (cause)• Prescriptive (recommend)
Growth Factors• Access to big data• Open access analytics &
crowdsourcing• Real-time data processing
• Complex event processing tools (CEP)• Parallel processing
Challenges (7V’s)• Veracity/validity of data• Legacy systems (unstructured &
real-time processing of data)
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
Act: Augmented Behavior
Types:
• Augmented intelligence drives informed action, while augmented behavior is an observable action in the real world.• Machine to machine
(M2M)
• Machine to human (M2H)
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
Intranet of Things: The Information Value Loop
• Augmented Behavior is the end of the information loop and results in an action:• Recommendation to a
provider from a BPG
• Text message to a diabetic patient to increase their insulin.
• Data Science Behavioral Science
Holdowsky, J., Mahto, M., Raynor, M. and Cotteleer, M. Inside the Intranet of Things. A primer on the technologies building
the IoT. Deloitte University Press.
1.Create(sensors)
2. Communicate(Networks)
3. Aggregate(AugmentedIntelligence)
4. Analyze(Integration)
5. Act(Augmented
Behavior)
Blood Glucose Monitoring
1. CreateBlood glucoselevels created
by a home monitor
2. CommunicateBlood glucose levels shared via
Bluetooth and the Intranet
3. AggregateBlood glucose levelstracked over time for
and individual or Population
5. ActPatient adjusts medications
and other factors(Augmented Behavior)
4. AnalyzeBlood glucose levelsare trended and care
plan created.(Augmented Intelligence)
Personal Analytics & Precision Medicine
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
Human Graphic Information System (GIS)
• Multiple layers of demographic, physiologic, anatomic, biologic and environmental data about a particular individual.
https://twitter.com/erictopol/status/449529091536863233
Topel, E. (2015). The Patient Will See You Now. Basic Books,New York.
Algorithmic Medicine
IBM Watson Health (cognitive computing-500G/S)
• Genomics
• Drug Discovery
• Value Based Care
• Patient Engagement
• Oncology
• Care Manager• https://www.youtube.com/watch?v=BYXIg1S7
nKk
Google DeepMind Health (open source)
• Founded in London (2010)
• Primary data source (NHS)
• Primarily unsupervised learning algorithms.
• Vision is to combine with neuroscience methods
https://www.youtube.com/watch?v=rXVoRyIGGhU
https://www.youtube.com/watch?v=wQU9wsFnO4k
Genomic – Precision Editing
Crisper-Cas9
• Molecular scissors target and snip out aberrant regions of genetic code, which can then be replaced with correct sequences.
http://www.yourgenome.org/facts/what-is-crispr-cas9
Face2Gene – Diagnostic Testing
• Deep learning algorithms classify photos into syndromes (for example Downs Syndrome).
• Software converts a patients photo into mathematical facial descriptors and compares them to database.
https://suite.face2gene.com/clinic-deep-phenotyping-of-genetic-disorder-dysmorphic-features/
Prevention: Asthma Attacks
Sensor Cluster:• Air quality
• Pollen
• Inhaler use
• Geo-location
• Breath nitric oxide
• Lung function –Smartphone app
• RR, Temp, O2 sat.
Wearable and Implantable Wireless Sensor Network Solutions for Healthcare Monitoring. Ashraf Darwish and Aboul Ella Hassanien : Accessed at Sensors website: http://www.mdpi.com/1424-8220/11/6/5561/htm
Prevention: Heart Failure Events
Sensor Cluster
• Beat to beat variability
• Fluid status
• Sleep quality
• Apneic spells
• Vital signs
• Lab tests (via smart phone app)
• Med. adherence (via digitized pills)
https://www.youtube.com/watch?v=-uTsMCvT7X8
Care Coordination & Advanced Monitoring Devices
Implantable Continuous Glucose Monitoring
•Provides education and monitoring
=>
http://www.medtronicdiabetes.com/treatment-and-products/continuous-glucose-monitoring
Health Coaching Tools: Patient Engagement and The Quantified Self
Wearable Computing
• Activity monitors
• Diet & weight loss monitors
• Sleep and mood
• HealthIt.gov
http://www.ted.com/talks/gary_wolf_the_quantified_self?language=en
http://www.healthit.gov/patients-families/stay-well#devices
Health Maintenance and Home Monitoring Devices• Home sensing
devices • Weight scale • BP monitor • Mattress monitors• Baby monitors• Spirometer
medication monitoring
• Pedometer http://video.brookstone.com/v/34618/fitbit-aria-wi-fi-smart-scale-l2giftsprice-l3gifts200/
http://www.youtube.com/watch?v=R-ypgw03sy0
Smart Homes: The Hospital Beds of the Future
• Video monitoring
• Continuous VS
• Gait sensors
• Smartphone symptom checkers
• Handheld Xray, lab
• Medication adherence dispensers
• Emergency alerts
• Remote temp control & security alarms
Image: http://www.iotforreal.com/lowes-enters-the-smart-home-market/2827
Safe Homeshttps://www.youtube.com/watch?v=C3FS8-Ka7SU
Augmented/Virtual Reality
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
Augmented and Virtual Reality
• VR uses software to simulate 3D images, sounds and sensations to isolate and surround you.
• AR overlays views of the physical world with fabricated images that engage users.
https://www.youtube.com/watch?v=N3ywcoqR4Co
https://www.youtube.com/watch?v=ysnVRWaYjLc
https://theglobalhealthnews.com/augmented-reality-makes-presence-felt-healthcare-industry/
General Purpose Machine Intelligence
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
When Computers Outperform HumansNarrow AI
When tasks are easily broken down into sequential, smaller arithmetical components, computers outperform the human brain.
Year Machine Beat Reigning Champion
• Backgammon (1979)
• Chess (1997)
• Checkers (2002)
• Scrabble (2002)
• Bridge (2002)
• Jeopardy (2010)Bostrom, N. (2014). Super Intelligence:Paths, Dangers, Strategies. OxfordUniversity Press.
Strong vs Weak Artificial Intelligence
Weak AI (Narrow)
• Non-sentient AI
• Focused on a limited number of narrow tasks.
• Most AI today is “weak”.• Siri – hybrid AI combining
several weak AI techniques and big data.
Strong AI• General intelligence.
• Ability to apply intelligence to any problem rather than on one specific problem.• Brain Emulation and Mind
Uploading - scanning mental state (including long-term memory and "self") of a particular brain substrate and copying it to a computer.
Machine Level Human Intelligence (MLHI)
When will human level machine intelligence (strong AI) be attained?
Year Percent Chance
2022 10%
2040 50%
2075 90%Bostrom, N. (2014). Super Intelligence:Paths, Dangers, Strategies. OxfordUniversity Press.
Super-intelligence: 50% chance 30 years after human level intelligence
MLHI
Smart Robots
https://www.artificialintelligenceonline.com/16353/machine-learning-is-at-the-very-peak-of-its-hype-cycle/
Smart Robots
• Sense• LIDAR (point cloud)• https://www.youtube.com/watch?v=5M1toZ
swwbA
• Think• Big data/machine
learning & AI• https://www.youtube.com/watch?v=O1Z
hWv84eWE&index=1&list=PLus_8yixy1UBIpIcT67_wmB_Pxd4g0KrC
• Act• Automotive Industry
• https://www.youtube.com/watch?v=-7xvqQeoA8c
https://www.youtube.com/watch?v=aWxF_ZXs-_o
The Next 10 Years….
Top Technology Trends
• Nanotechnology
• Information Tech
• Networks
• Neurotechnology
• Biotechnology
• Robotics
• Quantum technology
Canton, J. (2015). Future Smart: Managing the Game-Changing Trends That Will Transform Your World. DeCapo Press.
The New Future for Nurses Will Be Driven By:• Accelerated change
• Fast innovation
• Smart technology
• Predictive systems
• Connected markets
• Digital everything
• Mobile commerce
Canton, J. (2015). Future Smart: Managing the Game-Changing Trends That Will Transform Your World. DeCapo Press.
Image: http://www.accelerationwatch.com/
Emerging Roles
• Expanded scope of practice - APRN
• Digital Care Coordination
• Personalized medicine health consultant
• Population management teams
• Nurse data scientist
• Nurse entrepreneurs
https://thenpmom.wordpress.com/2012/01/01/the-future-of-nursing-a-nurse-practitioners-perspective/
Questions?Thomas R. Clancy, Phd, MBA, RN, FAAN
clanc027@umn,edu