• No results found

Sensoren voor de herkenning van activiteiten in zorgtoepassingen

N/A
N/A
Protected

Academic year: 2021

Share "Sensoren voor de herkenning van activiteiten in zorgtoepassingen"

Copied!
52
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Amsterdam University of Applied Sciences

Sensoren voor de herkenning van activiteiten in zorgtoepassingen

Kröse, B.J.A.; Englebienne, Gwenn; Hu, Ninghang; van Oosterhout, Tim; Kanis, A.M.

Publication date 2012

Document Version Final published version

Link to publication

Citation for published version (APA):

Kröse, B. J. A., Englebienne, G., Hu, N., van Oosterhout, T., & Kanis, A. M. (2012). Sensoren voor de herkenning van activiteiten in zorgtoepassingen. Hogeschool van Amsterdam.

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please contact the library:

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied Sciences), Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

Download date:27 Nov 2021

(2)

Sensoren voor de herkenning van activiteiten in zorgtoepassingen

Ben Kröse, Gwenn Englebienne, Ninghang Hu, Tim van Oosterhout, Hayley Hung, Marije Kanis

(3)

Noodzaak voor zorgtechnologie

(4)

Waarom willen we sensoren voor zorgtoepassingen?

We kunnen gezondheidstoestand afleiden:

Direct, met speciale sensoren

Heart rate

blood pressure

sugar level,

Indirect, door het activiteitenpatroon van de gebruiker te meten

(5)

Wat voor soort activiteiten meten we?

(6)

Wat voor sensoren zijn er?

(7)

‘Living lab’ aanpak

(8)

7

Activiteitenherkenning met eenvoudige sensoren

• Psychogeriatrische afdeling van Naarderheem

• Aanleunwoningen

(9)
(10)

9

(11)
(12)

Psychogeriatrische afdeling: Eenvoudige regels

(13)

De Flank: Aanleunwoningen ambient assisted living

12

(14)

Wat meten we?

7 apartments with 15 sensors each:

Bed

Diverse kitchen cabinets

Doors

Electrical appliances

Couch

Motion sensors

Toilet flush

13

(15)
(16)

CHI 2011 VANCOUVER MAY 7-12-2011 Marije Kanis

SENIOR CREATE-IT

Toilet flush sensor

(17)
(18)

Lange termijn data collectie

(19)

Hoe krijgen we zinvolle informatie?

Kenmerken bepalen

Gebaseerd op expertise van zorgverleners

Interview caregivers

Make features

Select best features

Statistische verwerking

ICA, PCA

Clustering

(20)

Visualizatie sensordata

19

# Badkamerpatroon:

(21)
(22)
(23)

Postprocessing

Indicatie aantal uren slaap per nacht (op basis van vuurpatronen van de bedsensor):

22

(24)

Vergelijk de sensor data met de metingen van professionele zorgverleners

Ergotherapeuten gebruiken verschillende methoden om gezondheid te meten:

Subjectief

modified KATZ ADL index

(self report on basic ADL, instrumental ADL)

Objectief

AMPS scale

(Physical performance: gait- speedtest 3-m measured walk, grip strenght-test Jamar Dynamometer)

23

(25)

Automatische

herkenning

(26)

Automatische activiteitenherkenning

Kunnen we meten wat voor algemene dagelijkse

levenstaken (ADL) iemand uitvoert?

(27)

26

Sensor patroon xt hangt af van de activiteit zt

De activiteit zt hangt af van de activiteit zt-1

Hidden Markov Model

(28)

Need a training set consisting of examples:

{

z1

,

x1

,

z2

,

x2

,... ,

zN

,

xN

}

Estimate parameters with ML methods

Het HMM moet getraind worden met

voorbeelden

(29)
(30)

Experiments: data

(31)

Activiteiten

(32)

Conditional random field

31

• Sensor pattern xt and activity zt are dependent

• activity zt and activity zt-1 are dependent

(33)

Conditional Random Field

Not a full probabilistic model

(more like a neural network)

Also training is needed

(34)

Hidden Semi-Markov Model

33

• A number of sensor patterns xt-n …xt depend on activity zt

• activity zt depends on activity zt-1

(35)

Resultaten

Gemiddeld 80% van de tijd correct

(36)
(37)

Training time

Conclusies: CRF iets beter, ten koste van

training tijd

(38)

Automatische herkenning

met videotoezicht

(39)

Valdetectie

Most common cause of injury with persons 55+

In the Netherlands annually 95.000 emergencies

Of which 43.000 in and around the home

1.3% fatal

Problem will increase with ageing population

(40)

Fall detection: existing solutions

Wearable accelerometers

Ambient detectors

Problems: not worn, restricted use

(41)

Oosterhout et al: 2D dynamics

(42)

Er zijn veel verschillende types camera

Infrarood

Fish-eye

Time of flight

Stereo

(43)

42

Fall lab

(44)

43

(45)

Oosterhout et al: 3D dynamics uit stereo

(46)

Better person modeling 2

(47)

Combineren 2 camera’s

(48)

Problemen

(49)

Problemen en uitdagingen

Goede datasets

Privacy issues

(50)

Ouderen vallen niet zo!

(51)

Privacy

Willen we camera’s

in onze woning?

(52)

Conclusies

Intelligente sensorsystemen lijken goed toepasbaar in de gezondheidszorg

Er is een tekort aan realistische datasets

Privacy issues moeten serieus genomen

worden

Referenties

GERELATEERDE DOCUMENTEN

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied

https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the University of Amsterdam and Amsterdam University of Applied