BioSigMA

Bio Signal Monitoring Application

Daniel Ferreira

and Gil Gonc¸alves

Increase Time, Matosinhos, Portugal

Faculty of Engineering of the University of Porto, Porto, Portugal

Keywords:

Medical Applications, Mobile Monitoring, Mobile Devices, Bio Signals.

Abstract:

The aging of the population causes an increased prevalence of chronic diseases, such as cardiovascular disease

(which is the leading cause of death in developed countries) and dementia. Because of the high morbidity and

mortality rates associated with this group of diseases, it is necessary to continuously monitor the vital signs of

people at risk. Nowadays this monitoring is carried out by a Holter monitor, which acquires electrocardiogram

data for a long period of time, so that it can be analyzed later. However, this is not a real-time monitoring.

There is another type of monitor, which stores the data and communicates with a remote server, allowing real-

time continuous monitoring. The latter system requires a speciﬁc platform and, as result, the patients have to

adapt to yet another device in their daily activities.

Information and communication technologies (ICT) have had a remarkable role in the management of health

care distribution and social work, and can be applied on daily monitoring of patients, providing a timely op-

portunity for medical staff to intervene. In the ICT ﬁeld we have witnessed the rise of smartphones as a gadget

with great mobility, connectivity and processing capacities. They are the ideal device to take patient moni-

toring to the next level, replacing the need for speciﬁc platforms for each type of monitoring, and facilitating

the daily lives of patients. This ability of smartphones becomes more and more apparent with the increasing

number in new medical applications that proﬁt from its characteristics.

Therefore, our goal is to create an application for smartphones which takes advantage of the portability and

processing capacities of smartphones to assess cardiac function, using bio signals captured by a device with

bluetooth interface and the sensors on the smartphone, and subsequent processing with a medical telemetry

system.

1 INTRODUCION

The aging of the population, caused by the in-

creased life expectancy and diminished number of

child births, in the occidental population, means we

have more and more elderly and less active people.

Due to the difﬁculties of some of the elder people

to live independently, there’s a necessity to maintain

a constant vigilance of their activities and their vital

signs. With aging come various problems: a great part

of this elder population have mobility issues, that can

be the cause of falls. 70% of all the deaths caused by

falls occur on the elder population (Edelberg, 2001);

also, the high prevalence of dementia in this popu-

lation can impair their orientation; a large number

of people suffer from cardiovascular disease, which

tends towards worsening with age. More than 30%

of the deaths in Portugal (Instituto Nacional de Es-

tat

ıstica, 2011) are caused by heart disease, that re-

quires a continuous monitoring. This monitoring can

be made using devices like the Holter monitor, that

acquires electrocardiogram data from a long period

of time to be evaluated by a doctor. However, this

device does not provide real-time monitoring, which

limits its usage to diagnosis. To address this gap, a

monitoring system can be used that consists of three

components: a network of sensors placed on users to

get their vital signs; a Platform for Personal Moni-

toring carried by the user that obtains the data from

the sensors and sends it to the third component; a re-

mote server where there will be a caregiver to view the

monitoring data. This type of monitoring is expensive

and requires users to use another device in their daily

living.

Our aim is to take advantage of the capabilities

such as processing power, communication interfaces

and sensors embedded in a smartphone, a device in-

creasingly common, in order to investigate its poten-

278

Ferreira D. and Gonçalves G..

BioSigMA - Bio Signal Monitoring Application.

DOI: 10.5220/0004323102780281

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2013), pages 278-281

ISBN: 978-989-8565-34-1

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

tial in the replacement of a speciﬁc Platform for Per-

sonal Monitoring.

This paper has 4 more sections: In section 2 we

present the State of the Art, including examples of

projects related to this work. In section 3 we describe

the functionalities of this Application and how they

were developed. In section 4 we present the results

obtained in the tests made to this Application. In sec-

tion 5 we draw conclusions about the work done in

this project, and make some considerations about fu-

ture work.

2 STATE OF THE ART

Related with this work we can enumerate projects

that have speciﬁc goals and projects with a more

broader scope. The Brickhouse Alert (BrickHouseAl-

ert, 2011) consists of a device that connects to a home

phone and receives data from a sensor wore by the

user. This sensor sends an alert to the device in case

of a detected fall. In this type of fall detection systems

there is another project, SmartFall (Lan et al., 2009),

that uses a modiﬁed crane to include accelerometers

and gyroscopes for detecting falls and communicates

alerts through Bluetooth. Both presented systems re-

quire speciﬁc hardware to be used by the user. To

address this and provide a more comfortable solution,

there are systems that use a daily device to implement

this functionalities. The PerFallD (Dai et al., 2010)

project uses the accelerometer in a smartphone for

detecting the falls, by recognizing patterns of force.

Today’s mobile platforms also present an increasing

number of mobile applications with medical use that

are able to read the cardiac rate through the smart-

phone camera (Azumio, 2011). There are also ap-

plications that implement geofencing capabilities, and

some of them, like the Reminders application of iOS,

are already embedded in the operating system. These

are all systems that take advantage of the functionali-

ties provided by the smartphones, but they differ from

BioSigMA in that they implement a system with only

a speciﬁc use. In the group of projects that have a

more broader scope we have the mobile monitoring

systems that mainly use the architecture presented in

1. The Heartronic project (Rocha et al., 2010) consists

of an array of sensors wore by the user that communi-

cate, using Bluetooth, with a smartphone. This smart-

phone sends the raw data to a remote server where it

will be processed, opposed to the presented solution

in this paper where the data is treated on the smart-

phone itself, before being sent to the server. Some

projects in this group have a speciﬁc population tar-

get, namely the elderly. Projects as the eCAALYX

Sensors

Platform for

Personal

Monitoring

Remote

Server

Figure 1: Overview of a real time monitoring system.

(Boulos et al., 2011), still in development, and iCare

(Lv et al., 2010), invest in the user interface of the ap-

plication, because it will be used by the elderly, and

needs to be simple and intuitive. The BioSigMA sys-

tem also has the same elderly target and implements a

simple user interface, but it also has advanced features

that can be accessed by capable users.

3 IMPLEMENTATION OF THE

BioSigMA App

To implement the Platform for Personal Monitoring

on a smartphone it was necessary to investigate which

development platform would bring more beneﬁts to

an application of this kind. We chose the Android

platform, because it is an open system, which pro-

vides freedom to develop and distribute applications.

For the purpose of this work it is important to have

this freedom so we have access to the smartphone

hardware without restrictions. Due to the nature of

this application, which requires a continuous moni-

toring, we used Background Services, that consists in

a process provided by the system that remains run-

ning in the background. This Service is the basis of

the application and manages the communication and

processing made by it.

3.1 Requirements

In order to implement a Platform for Personal Mon-

itoring, the BioSigMA App needed to fulﬁll the fol-

lowing requirements: Maintaining a continuous exe-

cution; Obtaining values from the monitoring sensors;

Storing the monitoring data; Conﬁguring the moni-

toring parameters; Managing the connected sensors;

Processing the received sensor data; Transferring the

data to the remote server; Reducing the data transfer

costs; Detection of falls; Location monitoring; Con-

ﬁguring notiﬁcation alert limits; Sending alert notiﬁ-

cations;

In the following subsections will be presented how

these functionalities were implemented in the Appli-

cation.

3.2 Sensor Communication

In order to communicate with the sensors available

BioSigMA-BioSignalMonitoringApplication

279

(Zephyr HxM, Zephyr BioHarness and a platform

for acquiring bio-signals developed by Jo

ao Oliveira

in his dissertation for the Master in Electrical and

Computers Engineering (Pedro and Oliveira, 2012)),

BluetoothSockets were used for the data transmission

through Bluetooth and Threads were used to make the

transmission asynchronous. Since each sensor has its

own transmission protocol, it was necessary to create

distinct classes to manage the transmission with each

type of sensor.

3.3 Bio Signal Processing

The data received from the sensors, that includes heart

rate, breathing rate and temperature values, is stored

locally, in a SQLite database, for local consulting.

The data is then veriﬁed against the caregiver’s de-

ﬁned limits, and in case it is beyond the limits an alert

is triggered locally, notifying the user, and remotely,

on the server. The data is sent to the server organized

in packages containing some seconds of information,

so there is no constant data connection with the server.

In case of alert the package with the last seconds of

information is sent immediately. The remote server

consists of a SOAP Web Service.

3.4 Geofencing

To implement the geofencing functionality in the ap-

plication, we used the location data tools provided by

the system. The Service registers itself as an handler

for location events, and when it receives a location

changed event it calculates the distance between the

current location and the location deﬁned by the care-

giver as the center point. If the distance is bigger than

the radius, also deﬁned by the caregiver, an alert is

triggered locally on the smartphone and remotely on

the server.

3.5 Activity

Using the accelerometer data, it is possible to know

the force acting on the smartphone, so we can know

if the user is moving, and evaluate the intensity of the

movement, allowing to correlate the monitoring data,

such as the heart rate, with the activity level, reducing

the triggering of false positives. The activity is eval-

uated using the mean of the last 30 values received

from the accelerometer so we can ease the spikes.

3.6 Fall Detection

The fall detection algorithm uses the data provided by

the accelerometer, like the activity algorithm, to de-

!#$"

!#%"

!#&"

!#'"

(#$"

(#%"

(#&"

(#'"

)!"

*!"

(+!"

$(!"

$,!"

))!"

)*!"

%+!"

+(!"

+,!"

&)!"

&*!"

,+!"

'(!"

',!"

*)!"

**!"

(!+!"

(((!"

((,!"

($)!"

($*!"

()+!"

(%(!"

(%,!"

(+)!"

(+*!"

(&+!"

(,(!"

(,,!"

(')!"

('*!"

(*+!"

$!(!"

$!,!"

$()!"

-../0/12345678"

9:;/"6;<8"

Figure 2: Force exerced on the smartphone in a fall.

tect a pattern as the one in the ﬁgure 2.

When a pattern of a minimum (bellow a certain

threshold) followed by a maximum (above a limit) is

detected, it’s possible that the user has fallen. To make

the detection more robust this force pattern is com-

plemented, in smartphones that include gyroscope, by

the orientation of the device that will be different be-

fore and after the fall. After a fall the user activity will

be monitored so we can know if the user recovered

from the fall, or if he/she remained immobilized. In

the case that he/she remained immobilized the smart-

phone will trigger an alarm. If the alarm is not can-

celled, a text message will be sent to the caregiver

with the location of the user.

4 INTEGRATION WITH KeepCare

AND RESULTS

This mobile application was integrated in an already

implemented solution (Sousa et al., 2012). This so-

lution consists of a mobile application that provides

the user interface for the BioSigMA platform, en-

abling the user to visualize his monitoring data, and

also conﬁgure the monitoring parameters, as the lim-

its for triggering alarms and the types of sensors con-

nected to the device. The KeepCare application con-

nects to a remote server (developed using the Wise

Framework, by FreedomGrow) that receives the data

from the smartphone and stores it, so the user is able

to see the history of the monitoring on a website. The

server also triggers alarms and shows notiﬁcations on

the website.

We tested the functionalities implemented using

as close to real-world examples as possible. The im-

plementation of the bio signal data communication

to the smartphone and from the smartphone to the

remote server was straightforward, so the tests were

BIODEVICES2013-InternationalConferenceonBiomedicalElectronicsandDevices

280

Figure 3: Monitoring data from Zephyr HXM.

simple.

In Figure 3 we can see the data as shown in the

mobile application. This data is received and shown

in real time. The alerts are also triggered in real time,

sending a notiﬁcation to the user. We also tested the

activity levels by doing walks and runs, so we could

adjust the thresholds at which we consider an activity

a low level activity or a high level activity. For test-

ing the fall detection, due to the nature of the tests,

we could not use real falls, so we simulated falls.

The simulation consisted on users walking slowly and

tripping, falling on a mattress with the back facing

up, down, left or right side. This simulated falls were

all detected, by the application. The difﬁculty resides

in not considering daily episodes as falls. Episodes

as walking downstairs have force patterns similar to

falls. So, if we have a smartphone with a gyroscope

we can determine the orientation before and after a

fall. In a fall the orientation differs within a value

close to 90

, while when walking downstairs the ori-

entation variation is smaller. When a gyroscope is not

available a timer is started when a fall is detected and,

if the timer is interrupted by the user, there will be no

notiﬁcation of a fall, if the timer runs out there will be

a fall notiﬁcation.

5 CONCLUSIONS AND FUTURE

WORK

With this work, a Platform for Personal Monitoring

was developed in a smartphone that, taking advan-

tage of the capabilities of this kind of device, imple-

ments functionalities that go beyond the transmission

of the vital signs from the sensors to a remote server.

Therefore, we can conclude that it is possible to use

this kind of generic device in substitution of a spe-

ciﬁc Monitoring Platform. In terms of future work

we will implement protocols for communication with

new types of sensors and study the cost/beneﬁts of

using smartphones with included gyroscope that pro-

vide better results in the fall detection algorithm.

REFERENCES

Azumio (2011). Instant heart rate. Retrieved February 2,

2012 from http://www.instantheartrate.com/.

Boulos, M. N. K., Wheeler, S., Tavares, C., and Jones,

R. (2011). How smartphones are changing the face

of mobile and participatory healthcare: an overview,

with example from eCAALYX. Biomedical engineer-

ing online, 10(1):24.

BrickHouseAlert (2011). Brickhouse alert. Retrieved June

4, 2012 from http://www.brickhousealert.com/.

Dai, J., Bai, X., Yang, Z., Shen, Z., and Xuan, D. (2010).

Perfalld: A pervasive fall detection system using mo-

bile phones. In Pervasive Computing and Communi-

cations Workshops (PERCOM Workshops), 2010 8th

IEEE International Conference on, pages 292 –297.

Edelberg, H. (2001). Falls and function: how to prevent

falls and injuries in patients with impaired mobility.

Geriatrics, 56(3):41.

Instituto Nacional de Estat

ıstica, I. (2011). Indi-

cadores sociais 2010. Retrieved from http://

www.ine.pt/ngt server/attachﬁleu.jsp?look parentBoui

=132500372&att display=n&att download=y.

Lan, M., Nahapetian, A., Vahdatpour, A., Au, L., Kaiser,

W., and Sarrafzadeh, M. (2009). SmartFall: An Au-

tomatic Fall Detection System Based on Subsequence

Matching for the SmartCane. Proceedings of the 4th

International ICST Conference on Body Area Net-

works.

Lv, Z., Xia, F., Wu, G., Yao, L., and Chen, Z. (2010). icare:

A mobile health monitoring system for the elderly.

IEEE-ACM International Conference on Green Com-

puting and Communications and International Con-

ference on Cyber, Physical and Social Computing,

0:699–705.

Pedro, J. a. and Oliveira, S. (2012). Desenvolvimento e

integrac¸

ao de sensores numa Plataforma para sistemas

de Monitorizac¸

ao Pessoais. Tese de Mestrado, Facul-

dade de Engenharia da Universidade do Porto.

Rocha, V., Borza, P., Correia, J., Goncalves, G., Puscas,

A., Seromenho, R., Mascioletti, A., Picano, A., Coco-

rada, S., and Carp, M. (2010). Wearable computing

for patients with coronary diseases: Gathering efforts

by comparing methods. International Conference on

Automation, Quality and Testing, Robotics, 2:1–9.

Sousa, R., Alves, M., and Gonc¸alves, G. (2012). Health

care management with KeepCare.

BioSigMA-BioSignalMonitoringApplication

281