1 uc berkeley uc davis uc merced uc santa cruz eldercare april 2006 ruzena bajcsy, mike eklund,...

27
1 UC BERKELEY UC DAVIS UC MERCED UC SANTA CRUZ Eldercare April 2006 Ruzena Bajcsy, Mike Eklund, Shankar Sastry, Steve Glaser Presentation by Ruzena Bajcsy

Post on 20-Dec-2015

217 views

Category:

Documents


0 download

TRANSCRIPT

1

UC BERKELEY

UC DAVIS

UC MERCED

UC SANTA CRUZ

Eldercare April 2006

Ruzena Bajcsy, Mike Eklund, Shankar Sastry, Steve Glaser

Presentation by Ruzena Bajcsy

2

White House Forum: Technologiesfor Successful Aging (10/4-5/00)

White House Forum: Technologiesfor Successful Aging (10/4-5/00)

Number of senior in U.S. estimated to increase from over 33 million today to:

» 53 million in 2020» 77 million in 2040

By 2023, the demographic profile for the nation will be similar to the profile in Florida today.

3

Parent Support Ratio:1950-2050Parent Support Ratio:1950-2050

0

5

10

15

20

25

30

1950 1990 2010 2030 2050

No. +85

As the number of elderly needing care increases, the number of potential caregivers decreases.

Today, 1 in 4 U.S.families care for anolder adult.

By 2005, nearly40% of U.S. workerswill be more concerned caring fora parent than a child.

Source: U. S. Census

Persons 85+ per100 people 50-64 years old

4

85+80-8475-7970-7465-6960-6455-5950-5445-4940-4435-3930-3425-2920-2415-1910-145-90-4

Age

1950(150,216,000)

1980(227,658,000)

2000(267,955,000)

2030(304,807,000)

Squaring the U.S. Population Pyramid1950-2030

Squaring the U.S. Population Pyramid1950-2030

5

Opportunities and ChallengesOpportunities and Challenges

Technologies that meet the challenges of aging will be increasingly valuable

We must identify collaborative, technology transfer, technology development and deployment opportunities for government, industry, and academia that help improve the independence, mobility, security, and health of aging U.S. citizens

We must examine potential opportunities and barriers and identify and prioritize recommendations both near and long term, including Grand Challenge class recommendations

6

Focus on Fall detectionFocus on Fall detection

While our agenda is much broader , that is we are interested in a comprehensive program of integrating sensory, wireless technology , embedded systems into an architecture that would facilitate monitoring the elderly population

Their medical records s well as their daily activities.

In this presentation we shall discuss only one aspect that is the Fall detection and alarm system.

7

Worldwide healthcare crisis is hereWorldwide healthcare crisis is here

- Every major world economy has health as biggest percentageEvery major world economy has health as biggest percentage- Nursing shortage in many parts of the worldNursing shortage in many parts of the world- South Korea and Japan technology infrastructureSouth Korea and Japan technology infrastructure

Trend #2Trend #2

8

Elder care is returning home againElder care is returning home againPoor Houses / AlmshousesPoor Houses / Almshouses

““pauper”pauper”

Insane AsylumInsane Asylum““inmate”inmate”

HospitalHospital““patient”patient”

Assisted LivingAssisted Living““resident”resident”

HomeHome““grandma”grandma”

Nursing HomeNursing Home““senior citizen”senior citizen”

HomeHome““grandma”grandma”

Only way to save costs but Only way to save costs but increase quality is home care.increase quality is home care.

Home care is fastest growing Home care is fastest growing segment of health industry.segment of health industry.

Trend #3Trend #3

9

Convergence is actually happeningConvergence is actually happening

• Everyday health through everyday devicesEveryday health through everyday devices• Everything is a touchpoint to everything elseEverything is a touchpoint to everything else• Every device has a chip, every chip as a radioEvery device has a chip, every chip as a radio

Trend #4Trend #4

10

1. Proactive Health Lab in various Institutions1. Proactive Health Lab in various Institutions

http://www.intel.com/research/prohealth/

11

SensorNet OverviewSensorNet Overview

Fall Detector

Berkeley Mote

RS-232

RS-232

E.g. Bluetooth Sender

E.g. Bluetooth Sender

Berkeley

MoteSensors

Zigbee

Sensors

Mobile Gateway

Home Health System

Mobile Phone

Integrated

Camera

Internetand/or

telephone

Berkeley

Mote s

Hospital

Terminal, WLAN

12

Security and Privacy: WirelessSecurity and Privacy: Wireless

Bluetooth has built in (and evolving) security. » There are three modes of security for Bluetooth access

between two devices.• Security Mode 1: non-secure• Security Mode 2: service level enforced security• Security Mode 3: link level enforced security

ZigBee (802.15.4) security includes methods for » key establishment, » key transport, » frame protection, and » device management.

13

Wireless Network: CapabilitiesWireless Network: Capabilities

E.g., Wireless video consultation» Uses local Bluetooth, 802.11 network and/or ethernet» Monitor sensors, medical devices, etc.

Fall DetectorSensorsUser, or Health Care

Professional , etc

Server

Health Care Professional

14

Fall DetectorFall Detector

Features:» 3-axis, ±10 g accelerometers» on board GPS» Battery powered» RS-232 connection» C programmable» 80 Hz sampling» 4 hours recording,

or continuous streaming Functionality:

» Record or stream accelerometer data for testing» Embedded fall detection algorithms» Connect to Bluetooth radio

• Bluegiga device• Initiate a connection with a control device, i.e. the home or

mobile gateway

15

Basic Signal ProcessingBasic Signal Processing

Raw XYZ

Calibration

Low Pass 1 Low Pass 2

CoordinateTransformation

Align toWorld XYZ

PoseRecovery

MotionClassification

Three steps:» Calibration» Filtering

• Remove transients and noise

» Coordinate Transformation

• For pose recovery and orientation

16

Pose Recovery (of device):Filtering & TransformationPose Recovery (of device):Filtering & Transformation

Gravity is always present» Use low pass filters to reveal

the Earth frame of reference Transform to Spherical

Coordinates to reveal gravity and orientation

( )

( )

2 2 2

2 2

arctan

arctan

R x y z

y x

x y z

θ

φ

= + +

=

= +

XYZ Data, filtered and not

Spherical Data, filtered and not

Angle off vertical (rad) Φ

Gravity

Rotation θ

17

The Experiment in SonomaThe Experiment in Sonoma

We have conducted 3 experiments each with 3-5 people wearing our device for approximately two hours.

First we have instructed the persons to perform repeated sit-down stand-up, walk on stairs, walk on floor, each for 5 times during which we recorded the acceleration.

Second we left the person to do normal activities while wearing the sensor for approximately 1 ½ hours.

18

Activities of InterestsActivities of Interests

Activities of Daily Living» Walking» Sitting» Standing» Other normal activities

Falls» Categorizing the types and severities

19

Example ResultsExample Results

The recorded data from accelerometer based fall sensor is analyzed and replayed with this Matlab program.

Examples» an elderly woman walking

with a walker (top right), » a woman sitting down (top

left), » a woman standing up

(bottom right)» a man stretching his arm

(bottom left)

20

Example of Identifying ADLs: SittingExample of Identifying ADLs: Sitting

Heavy lines represent average results in X, Y, Z

21

Example of Identifying ADLs: StandingExample of Identifying ADLs: Standing

Heavy lines represent average results in X, Y, Z

22

Example of Identifying ADLs: WalkingExample of Identifying ADLs: Walking

Heavy lines represent average results in X, Y, Z

23

Comparing: falling and sittingComparing: falling and sitting

Falling (Trained Judo-ist)

0 100 200 300-2

0

2

4

6X Accel

0 100 200 300-2

-1

0

1

2Y Accel

0 100 200 300-3

-2

-1

0

1

2

3Z Accel

0 100 200 3000

2

4

6

8Norm

Sitting (Septuagenarian)

0 200 400 600 800-2

0

2

4

6X Accel

0 200 400 600 800-2

-1

0

1

2Y Accel

0 200 400 600 800-3

-2

-1

0

1

2

3Z Accel

0 200 400 600 8000

2

4

6

8Norm

24

Lessons learnedLessons learned

While the signal for sit-down and stand-up a clear distinguishable characteristics, it is also clear that there are substantial differences from person to person that need to be accounted in order to avoid false positives.

This suggest that the data must be normalized with respect to the weight and height of the person.

25

But what about in-between?But what about in-between?

False positives» Unnecessary alerts, worries» At the least very bothersome

False negatives» Critical problem

Solution must be very robust:» Good algorithms and devices» User interaction» Sensor fusion» Reliability

26

Respect for PrivacyRespect for Privacy

As we interviewed the participants we learned that if they indeed would have fallen, they wish to alert» first the neighbors » then the children/relatives » and only later the 911

27

Q&AQ&A