Post on 14-Jun-2015
description
A M B R E , A R N A U D , H A R S H A , M A H E N , S H A H R O K H
WEARABLE ELECTRONICS IN HEALTH CARE APPLICATIONS
AGENDA
I. Introduction
II. Data Collection
III. Data Processing
IV. Data Display
V. Impact on Healthcare
VI. Is success of wearables possible ?
VII. Conclusion
I. INTRODUCTION
Wearables are small electronic devices, often consisting of one or more sensors and having computational capability. They play an important role in healthcare monitoring, analyzing and even healing.
WEARABLE ELECTRONICS BY SECTOR
FOCUSING ON HEALTH CARE
We
ara
ble
Ele
ctr
on
ic D
evi
ce
s
Hand Worn
Smart Watch
Wrist Wear Finger Wear
Head Worn
Smart Glasses
HMD / HUD
Body Worn
Smart Textile
Wearable Patches
Foot and Arm Wear
Apple Samsung Jawbone
Google Optinvent
OMsignal Intel
WEARABLE ELECTRONICS BY USE
§ Fitness and health tracker § IntelligentM Bracelet (how well you wash your hand)
§ Wearable computers § Amon
§ Watch § Pebble Smartwatch § Martian Notifier Smartwatch § Apple watch § Samsung gear
§ Wristband § MIT Wristband § The Tactilu Bracelet
HAND WORN
Hand Worn
Watch
Wearable computer
Wristband
§ Smart glasses § Google Glasses § Vuzix § Optinvent ORA § Buhel
§ Medical headsets (EEG)
§ Breathing masks
§ Brain-sensing headband (Muse-InteraXon)
§ Communication helmets § O.R.B
HEAD WORN
BODY WORN
§ Smart textile § Smart T-Shirt with integrated sensors (fitness trackers) § Smart armband (Myo) § Safety baby worn blanket (Philips)
§ Foot and Arm wear § RunScribe
§ Wearable Patches
§ E-skin
COMPONENTS OF WEARABLES
Sensors Inertial sensors Biosensors
Other sensors
(Haptics…)
Connectivity Bluetooth WiFi GPS
Battery Conventional Flexible battery
Energy harvesting
module
Interfaces Speech recognition
Haptics / Touch
recognition
Gesture recognition
Non-invasive Interfaces
Materials /Algorithms
Electronic textiles and
joints
Flexible displays
Accurate interpretation of measured
data
RATES OF IMPROVEMENT
Source: http://www.newelectronics.co.uk/electronics-blogs/powering-wearables-and-giving-batteries-a-better-life/64664/
Next generation of Wearable devices
Batteries are the bottleneck Change in the architecture and power usage of ICs to make them more efficient
DATA COLLECTING
Inertial Sensors
ü To monitor body movements
Bio-Sensors
ü To monitor heart rate
ü Cholesterol
ü Sweat
Haptics
ü To enhance touch experience
INERTIAL SENSORS TO TRACK BODY MOVEMENT
ü Continuous real-time
data recording
ü Accurate
ü Body angles
ü Angular acceleration
Accelerometers and gyros
INERTIAL SENSORS TO TRACK BODY MOVEMENT
Integrated Motion tracking
Remote patient monitoring
Patient’s motion data
Doctors and
Physician
InvenSense’s Motion Tracking device
² 6 axis (3-axis accelerometer)
+ (3-axis gyroscope)
² 9 axis (additional 3-axis ecompass)
Source: http://www.invensense.com/mems/wearablesensors.html
INERTIAL SENSORS TO TRACK BODY MOVEMENT
Nike FuelBand FitBits
Basis Jawbone
OPTICAL SENSORS TO MONITOR HEART RATE
ü Acceptable accuracy for over than 15 min use
ü Commonly used in wrist bands
Wearble Sensors, ISBN: 978-0-12-418662-0
good accuracy
BIO-SENSORS TO MEASURE CHOLESTEROL
Electrochemical
ü Electrochemical are considered
to be the most important
cholesterol biosensor
ü Based on enzymatic catalysis of
a reaction
ü Low response time
ü High sensitivity
ü Low cost and low power required
Optical
ü Employs an optical fiber as a
platform for the biological
recognition element
ü Involves diffusion of analytes
ü Higher response time
ü Good sensitivity
ü High cost and high power required
PERFORMANCE OF ELECTROCHEMICAL VS OPTICAL
Electrochemical sensors seem to dominate pertaining the performance
Source: http://www.slideshare.net/Funk98/cholesterol-bio-sensors-getter-better-fast
NON-INVASIVE ASSISTIVE INTERFACES
ü Brain computer interface vs. Tongue control
interface
ü Tongue computer interface might be better ?
Mean responsibility of correct choices
Information transfer rate (bits per min)
Source: Wearable Sensors, ISBN: 978-0-12-418662-0
Type Number of Commands
Response Time (s)
IRT (Bits/min)
EEG-BCI 2 - 4 3 - 4 25
TTK-TCI* 9 3,5 40
TCI*-1 5 2,4 58
TCI*-2 6 1 95
Table: Comparison between the Tongue Drive System and other BCIs/TCIs*
* TCI Tongue Computer Interface
SWEAT SENSOR
ü Used in wearable textiles
ü Considerable
improvements required
Graph: Textile humidity sensor (upper left) and its calibration curve compared to a commercial humidity sensor
Source: Wearable Sensors, ISBN: 978-0-12-418662-0
HAPTICS TO ENHANCE TOUCH EXPERIENCE
ü Enables virtual reality
ü Weight illusions based on fingertip deformation
ü Sensorimotor enhancer improves tactile sensitivity in human fingertips
Source: Wearable Sensors, ISBN: 978-0-12-418662-0
Graph: Desired and measured eccentricity
DATA PROCESSING
CPUs and Processors Algorithms
Data Processing
DATA PROCESSING
Internal Processing
ü Data is processed within the wearable
ü Higher battery consumption
ü Efficient algorithms required
External Processing
ü Data sent to another device or cloud
ü Data processing on another device
ü Could use higher computational capabilities
Apple processor unit for healthcare and fitness data processing:
ü Embedded accelerometer, gyroscope and compass
ü Online process of motion data
ü Analysis of motion-related healthcare problems
ü Tested in IPhone 5 and will be used in Apple watch
PROCESSOR UNITS
PROCESSOR TRENDS
Source: “Wearable biosensing: signal processing and communication architectures issues” P. Cleka, R. Vetter, J. Telecom. Info. Tech, 2005
ü Performance
ü Power consumption
ü flexibility
Past trend Future trend
FIRST GEN OF WEARABLE PROCESSOR
Ineda systems Hierarchical CPU
ü Devised for wearable
electronics
ü Nano: always on
ü Low power consumption
ü Support more
sophisticated display
and input requirements
ALGORITHMS
Source: “Wearable biosensing: signal processing and communication architectures issues” P. Cleka, R. Vetter, J. Telecom. Info. Tech, 2005
Noise reduction !
ALGORITHMS FOR SPEECH RECOGNITION
ü Reasonable accuracy
ü Better algorithms are being developed
Wearble Sensors, ISBN: 978-0-12-418662-0
ALGORITHMS FOR GESTURE RECOGNITION
ü Good precision for higher samples
ü Well established algorithms are currently available
Source: Wearable Sensors, ISBN: 978-0-12-418662-0
ALGORITHMS FOR EEG AND ECG
Algorithms for EEG and ECG
ü High sensitivity
ü Accurate
ü Power performance
ü Detection rate
Source: Wearable Sensors, ISBN: 978-0-12-418662-0
CONNECTIVITY
Bluetooth v4.0 includes Bluetooth low energy marketed as “Bluetooth smart”
0
10
20
30
40
50
60
70
80
2012 2013 2014 2015 2016 2017 2018
% w
ith B
lue
too
th lo
w p
ow
er c
hip
s
all wireless sports and fitness monitoring devices (according to IMS research)
>70%
Source: http://www.nordicsemi.com/eng/News/ULP-Wireless-Update/Health-improvements-by-the-numbers
INTERNAL VS. EXTERNAL PROCESSING
Internal
ü Devices with low computational requirements
ü High performance to size ratio of processors
ü Connectivity is poor ü Processing is essential to display
results
Ex: Smart Watch
External
ü Devices with higher computational requirements
ü Existing processor capabilities are enough
ü Connectivity is not a bottleneck
ü Internal processing is not essential
Ex: EEG devices
HYBRID devices ?
DATA DISPLAY
Internal Display
ü Data displayed in the device or projected
somewhere
ü Flexible display and electronics desired
ü Larger wearables
External Display
ü Data displayed in another device (E.g. Phones,
tabs)
ü Existing display devices are sufficient
ü Smaller wearables
FLEXIBLE DISPLAYS
ü Better materials need to be discovered
ü Low stiffness, low thickness, better resolution are desired
ü Production costs are falling
http://www.displaysearch.com/cps/rde/xchg/displaysearch/hs.xsl/140716_amoled_mobile_phone_panel_costs_expected_to_fall_below_lcd.asp
FLEXIBLE ELECTRONICS
ü Flexible electronics would be very important
ü Strain vs. performance of transistor is an indicator
Wearble Sensors, ISBN: 978-0-12-418662-0
INTERNAL VS. EXTERNAL DISPLAY
Internal Display
ü Devices with minimum information to be displayed
ü Devices that can project data
ü Connectivity is poor
ü Larger size is desirable
ü Improvements in flexible, thin display systems
ü Improvements in flexible electronics also necessary
Ex: Wrist Bands
External Display
ü Devices that need detailed
analysis of data
ü Connectivity is good
ü Existing display systems are
sufficient
ü Smaller size is desirable
Ex: Medical Devices
Hybrid systems likely to be used mostly!
IMPACT ON HEALTHCARE
q Fitness tracking and improvement
q Management of hospital organization
q Personal drug dosage tracking
q Tele-medicine
q Rehabilitation
q Healthcare Big Data
FITNESS TRACKING AND IMPROVEMENT
Hand-worn and body-
worn
• Jawbone, Apple watch, Polo Tech Shirt
Tracking • Personal physiological and
biological parameters, activity and performance
Data collected
• Heart rate, stress, obesity, sleep, calories, 02 saturation, blood pressure
Virtual coaching • Apps
RUNSCRIBE
A small 9-axis sensor
Uploaded via Bluetooth to your devices
Attached to the heel of any shoe
Data stored locally
Measures 13 kinematic
metrics pace Impact
Gs Braking
Gs Pronation excursion
Pronation velocity …
Weight 15 g
MILLION KG CHALLENGE
Application: Million Kg Challenge
ü 80,000 signed up
ü 42,000 pledging to lose weight
ü 6,000 lost overall 20,000 Kg in 6 months
Apps § iDAT § MyFitnessPal § RunKeeper § LoseIt
Wearable fitness trackers § Jawbone Up24 § Withings Pulse O2 § Samsung Gear Fit § Nike+ FuelBand SE § Garmin Forerunner 15 § Apple Watch (available 2015)
MANAGEMENT OF HOSPITAL ORGANIZATION
Collecting
ü Continuous tracking of physiological data ü Hand, head & body-worn ü Sterility
Processing
ü Identify priority patients according to their needs ü Better diagnosis
Display
ü Efficient staffing of nurses and doctors ü Avoid useless displacement of staff ü Efficient drug and equipment management ü Assisting doctors in operating rooms
basis
PERSONAL DRUG DOSAGE TRACKING
Wearable drug reminding devices
ü Haptic, visual or sound drug reminder
ü Drug taking devices for the elderly
Wearable tracking and healing devices
ü Insulin monitoring – direct injection by the wearable
ü Baby care – fever, pain, antibiotics…
Insulin Nano-pump with MEMS
Apps to remind you when to take your pills
Sproutling wearable baby monitor
TELE-MEDICINE
Local and International applications
ü Health information technology
ü Distant and early diagnosis
ü Emergency tele-medicine
ü Health assistance to third world countries
ü Tele-medicine for soldiers on the battle field
Tele-medical services
ü Tele-pharmacy, Healthcare delivery
ü Tele-radiology, tele-cardiology
ü Tele-psychiatry
ü Tele-nursing
REHABILITATION
Handicapped people
ü Physical activity tracking to evaluate improvement –
better identification of physical weaknesses
ü Optimized HCI control – Voice control, non invasive
tongue control…
ü Sensors to compensate the loss of sight or speaking
ü eLEGS to help paraplegics to walk
The Elderly
ü Posture tracking to avoid falling
ü Retirement houses: better accommodation to
wearable computing
REHABILITATION: NUS PROJECT
Indoor guiding device for blind people
Large pool of data about health population
HEALTHCARE BIG DATA
Aggregate data from wearables with other health
information
Bring together people with a common interest
such as weight loss
Create a community
Build engagement and compile information
Complete and essentially real-time data to treat and manage the health of individual patients
Opens up Entrepreneurial Opportunities www.healthcaredatasolutions.com
IS SUCCESS OF WEARABLES POSSIBLE?
q Forecasts on Wearables and Healthcare
q What Reality in Healthcare ?
q The Right Time
q Success of Wearables as Interconnected Devices
Head-worn
Body-worn
Hand-worn
FORECASTS ON WEARABLES
2014
$3-5 billion
2018
$30 billion
2024
$94 billion
35% CAGR 2014-2024
Global market of wearable electronics
Wearable devices
2013
14 million
2018
500 million
Source: Deloitte, Visiongain, BIS Research, IHS
FORECASTS ON HEALTHCARE
Source: MaRS Market Insights, March 2014, Wearable Tech: Leveraging Canadian Innovation to Improve Health
Worldwide Public and Private Health Expenditure
ü Estimated at $7.3 trillion in 2012
ü 7% estimated annual growth in the next decade ü Personal care consumption: $1.7 trillion in 2012 should
reach $2.2 trillion in 2017
ü Tele-health patients to near 2 million by 2018
Healthcare wearables
ü ~30% of wearables market value: already $1.1 billion in
2014
WHICH REALITY IN HEALTHCARE ?
Dreams
• Instant access 24/7
• High quality healthcare
• Empowerment
• My health data
• Coordinated and coherent
• One point of contact
• Moving health care to me
Reality
• Gatekeepers
• Budget driven
• Didactive and controlling
• Unavailable health data
• Fragmented
• Falling between sectors
• Patients move to healthcare
Source: Moving healthcare to your fingertips: Klaus Phanareth at TEDxCopenhagen 2012 https://www.youtube.com/watch?v=_cGm_wJbyhU
THE RIGHT TIME
Time Magazine, September 22, 2014 ü Improved customer value proposition
ü Numerous features enabled by scaled sensors and displays
ü Established adoption of smartphones, tablets – products close to wearables
ü Rising life expectancy, more seniors and chronic diseases
ü Decrease the length of hospital stays
ü Long-term care
INTERCONNECTION OF WEARABLES
Jawbone RunScribe Apple Watch
Polo Tech Shirt
Google Glass
Cloud
Computers, smartphones, tablets and the Cloud
Apps
Wearables
INTERCONNECTION OF WEARABLES
Jawbone RunScribe Apple Watch
Polo Tech Shirt
Google Glass
Wearables
Most successful wearables will be
ü ‘Plug and play’ compatible with
all wearables
ü Bounded to an ergonomic app
ü Connected to the Cloud to save
and manage data
CONCLUSION
ü Real and meaningful purpose for Wearable Electronics
ü Collection of all physiological and biological data
ü Worldwide health data management
ü Breakthrough applications will emerge
A BREAKTHROUGH APPLICATION: BREAST TISSUE SCREENING
A real issue
§ Each year, 1,000,000 women are diagnosed with breast cancer
§ More than 400,000 die
§ 1 in 8 women contracts some form of breast cancer
Early detection is still the cornerstone
3 years 12 years
⇒ The First Warning Systems' Breast Tissue Screening Bra to assist in the breast self exam (BSE)
§ Painless
§ Noninvasive
§ Highly accurate
As easy as wearing sports bra
3 clinical trials 90% + of accuracy
Predicts tissue abnormalities
Collect data and send to the Internet
Process with sophisticated
algorithms and display to the user
A BREAKTHROUGH APPLICATION: BREAST TISSUE SCREENING
Thank you !
Q & A