wearable sensors and feedback system to improve breathing

Wearable sensors and feedback system to improve breathing technique

UNIVERSITY COLLEGE DUBLIN � DUBLIN CITY UNIVERSITY � TYNDALL NATIONAL INSTITUTE

Shirley Coyle 1, Edmond Mitchell 2, Noel O’Connor 2, TomásWard3 and Dermot Diamond 2

1 CLARITY Centre for Sensor Web Technologies, NCSR, D ublin City University, Dublin 9

2 CLARITY Centre for Sensor Web Technologies, CDVP, D ublin City University, Dublin 9

3Biomedical Engineering Group, Dept. of Electronic E ngineering, National University of Ireland Maynooth

Overview

• Introduction

• Sensors

• Data acquisition and processing


• Data acquisition and processing

• User feedback

• Conclusions

Importance of Breathing

Breathing – naturally occuring

Unique system– both voluntary and involuntaryWe can influence the involuntary autonomic nervous system using our


nervous system using our voluntary breath

By breathing in a slow, deep and regular manner, the heartbeat become smooth and regular, blood pressure normalizes, stress hormones drop, and muscles relax.

Importance of breathing technique

Sports performance

• Use of full lung capacity to maximise oxygen delivery to muscles

• Use of breathing techniques to calm and focus, e.g before kicking a penalty in soccer or a serve in tennis


Clinical applications

• Chronic Obstructive Pulmonary Disease (COPD)

• Anxiety treatment

• Cystic fibrosis

• Respiratory rehabilitation

Breathing exercises for patients


Kigin C., Breathing Exercises for the Medical Patien t: The Art and the Science, Physical Therapy/Volume 70, Number 11, November 1990

Aim – to develop a feedback system for patients to improve their breathing technique, by monitoring thoracic and abd ominal movements. Give feedback to the user graphically

Sensor must beComfortableRobustStraightforward to use

Wearable sensor/”Smart garment”

Breathing monitoring system –system requirements


Wearable sensor/”Smart garment”

Feedback/Application must: Grab users attentionHave a simple interfaceFocus user for the full duration of the programEncourage user to correct their breathing rate

and low cost, easy to install on computer systems

Breathing monitoring system

Wearable Sensor Microcontroller

and


Wearable Sensor - detect body

movement due to breathing

andWireless connection

Data processing and

User feedback

Fabric stretch sensors to measure body movements

Carbon black

++Elastomer

Cure Piezo-resistive sensor


Wacker ELASTOSIL® LR 3163 A/BLiquid Silicone Rubbers (LSR)

Screen printed onto fabric

Carbon elastomer stretch sensors

300

350

400

4502mm

4mm

7mm

11mm

Stretch sensor - Carbon loaded rubber screen printed onto lycra fabric. Elongation causes increase in resistan ce

2

47

11

10


0

50

100

150

200

250

0 1 2 3 4 5 6 7

Elongation (cm)

kOhm Sensor widths

2mm, 4mm, 7mm, 11mm, (length 10cm)

Average Gauge factor l(R-Ro)/(R(l-lo) = 1.6

Carbon elastomer stretch sensors

Sensor lengths 20cm, 14cm, 10cm 300

400

500

600

700

kOhm

20cm

14cm

10cm

20

14

10


20cm, 14cm, 10cm (width 5mm)

0

100

200

0 2 4 6 8 10

Elongation (cm)GF

20cm 2.17049414cm 1.00624610cm 0.486631

Sensors – “Smart shirt”Fabric stretch sensors monitor the expansion and contraction of the ribcage and abdomen during breathing.

4 Carbon-Elastomer(CE) sensors (piezo -resistive) are screen -printed Micro-

CE Sensor


(piezo -resistive) are screen -printed onto the front of the t-shirt

Sensors connected using conductive stainless steel thread.

Resistor leads are embroidered

Sew-in micro-controller Conductive thread

Micro-controller

Breathing monitoring t-shirt

Sensor 1

Sensor 2

Investigation of thoraco-abdominal movement using CE sensors


-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

0 5 10 15 20 25 30 35 40

Time(s)

Chest

Abdomen

Breathing signal (Normalised data )

Abdominal breathingThoracic breathing


SensorArduino

microcontroller User feedback

Arduino- open-source electronics prototyping platform


- open-source electronics prototyping platform

Arduino Pro MiniSize: 1.7cm x 3.4cmAnalog Input Pins : 6Digital I/O Pins :14 Microcontroller: ATmega168 or ATmega328

Arduino LilypadSewable microcontroller

Bluetooth connectionBluesmiRFWireless serial cable replacementTransmits any serial stream from 9600 to 115200bps


SensorArduino microcontroller

FlashApplication

The name Physput is derived from Phys iological input


Allows a user to emulate standard computer input using non-standard input.

Primarily designed to facilitate the design of alternative input devices.

Reads data from the serial port and maps this data to standard input e.g. mouse movements and key presses.

Version 1 written by Edmond Mitchell

PhysputWritten by Edmond Mitchell


SensorArduino microcontroller

FlashApplication

Most popular method for adding animation and interactivity to Web Pages.


Pages.

Can create rich Internet applications

Available free for web browsers such as a Internet Explorer and MozillaFirefox

99.3% 1 of all Internet desktop users have the Flash Player installed

User feedback interface

Start of breathing exercise

02468

101214161820

0 50 100 150 200 250 300 350Bre

athi

ng R

ate

(bre

aths

/min

)


0 50 100 150 200 250 300 350Bre

athi

ng R

ate

(bre

aths

/min

)

Time(s)

Sensormedics Vmax

Breathing patterns measured using the textile sensor and gold standard Sensormedics Vmax

Feedback application encourages constant breathing rate in syncronisation with the avatar

Deep breathing exercise

115 120 125 130 135 140 145 150 155 160-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5T

hora

co-a

bdom

inal

mov

emen

t

abdomenchest

Before

breathing

exercise


115 120 125 130 135 140 145 150 155 160

Time(s)

265 270 275 280 285 290 295 300 305 310-1

-0.5

0

0.5

1

1.5

2

Time(s)

Tho

raco

-abd

omin

al m

ovem

ent

chestabdomen

During

breathing

exercise

User feedback

0 5 10 15 20 25 30 35 40 45 50-3

-2

-1

0

1

2

3

Time(s)

Tho

raci

c m

ovem

ent

Measured

Reference

Grade A

R~90%


After performing the exercise the user is presented with a breathing efficiency grade.

The grade is calculated by cross correlating the users breathing pattern with a reference breathing pattern embedded in the program

0 10 20 30 40 50-3

-2

-1

0

1

2

Time(s)

Tho

raci

c m

ovem

ent

Measured

Reference

Grade F

R~90%

R~40%

User feedback – multiple sensor input


Real-time sensor signal versus computer generated accurate signal

The avatar encourages the user to perform diaphragmatic breathing.

The user aims to synchronise their breathing signal with the avatar

Real time feedback is given

Score given at the end of the exercise

Two sensors - placed at chest and abdomen

Breathing trainer


Conclusions/future directions

•Developed wearable wireless system to measure breat hing patterns, using textile sensors

•Developed user-friendly interface to help users imp rove their breathing

•Low cost, accessible system


•Low cost, accessible system

•Next stage to consult the target user groups for fe edback –Physiotherapists and Occupational therapists

•Clinical trials

Thanks to…- Edmond Mitchell , Centre for Digital Video Processing, CLARITY Centre for sensor Web Technologies, DCU

- Dr. Tomás Ward , Biomedical Engineering Group, National University of Ireland, Maynooth

- Stroke Rehab Unit, Adelaide and Meath Hospital, Tallaght


- Gregory May, Dept. of Human Health and Performance, DCU

- Prof. Noel O’Connor, Prof Niall Moyna, Prof. Dermot Diamond , CLARITY Centre for sensor Web Technologies, DCU

- This work is supported by Science Foundation Ireland under grant 07/CE/I1147

wearable sensors and feedback system to improve breathing

Documents