W2M2: WIRELESS WEARABLE MODULAR MONITOR

A Multifunctional Monitoring System for Rehabilitation

Antonio J. Salazar

1,2

, Ana S. Silva

1,2

, Claudia Silva

, Carla M. Borges

1,2

Miguel V. Correia

1,2

, Rubim S. Santos

and João P. Vilas-Boas

Instituto de Engenharia de Sistemas e Computadores do Porto (INESC Porto), R. Dr. Roberto Frias 378, Porto, Portugal

Faculdade de Engenharia da Universidade do Porto (FEUP), R. Dr. Roberto Frias s/n, Porto, Portugal

Centro de Estudos do Movimento e Actividade Humana (ESTSP-IPP), R. Valente Perfeito 322, V. N. Gaia, Portugal

CIFI2D, Faculdade de Desporto da Universidade do Porto, Rua Dr. Plácido Costa 91, Porto, Portugal

Keywords: Wearable, Rehabilitation, Modular monitoring system, Biosignal/inertial data logging.

Abstract: Wearable/portable biometric/physiological monitoring devices are rapidly becoming a recognized

alternative in medicine, rehabilitation and sports. Developments in sensors, energy harvesting, embedded

technology, smart textile, to mention a few, are driving the field to more seamless and complex solutions,

sometimes part of pervasive strategies for activity monitoring. Additionally, the number of sensors forming

part of wearable solutions seems to be incrementing thanks to miniaturization and lowering components

cost. Consequently medical and rehabilitation protocols and standards are undergoing the slow process

required for adaptation to such emerging trends. This article presents a simple, modular, low-cost, wearable

device originally intended for rehabilitation data gathering. Such device was based on commercially

available components which can be assembled and managed by physicians, therapist and other healthcare

personnel through a proposed platform. The objective is the familiarization and even active inclusion of

healthcare personnel in the technological development process and, more importantly, the incorporation of

electronic data acquisition in their procedures.

1 INTRODUCTION

Rehabilitation, healthcare and sports share a natural

affinity, especially in this age when healthcare

management seems to focus more on healthy

lifestyles and prevention. Such affinity is not limited

to the upkeep of a healthy condition through

exercise or the recuperation of an injury through

rehabilitation, but also at a technological level.

Nowadays, technology is providing new tools to all

three fields and entering within the individual’s

home at times. Wireless sensor networks,

implantable devices, textile integrated circuitry and

other achievements have allowed for designs that

require less energy and remain on site for longer

periods of time. Some systems are appearing now on

the market capable of human movement analysis and

for physiotherapy assistance. However, their

institutional oriented approach makes them

impractical for widespread usage due to highly

technical learning curves and/or required

accommodations size and electrical requirements;

not to mention prohibitive costs.

An interdisciplinary team was formed in order to

design and implement a simple and low-cost data

gathering platform, originally focused for the

development of quantifiable rehabilitation progress

methodologies. Commercially available components

were considered with the premise of the final

solution being adaptable and maintainable by non-

technical proficient persons; thus, high available

components and user-friendly solutions were

preferred.

1.1 Rehabilitation Protocols and

Procedures

The fields of physiotherapy and rehabilitation are

rapidly developing areas of health science that

covers a wide spectrum of multidisciplinary

intervention with several research and clinical

subspecialties. Rehabilitation in its broaden sense is

becoming a cornerstone of interest for both policy

makers and service providers within health and

213

J. Salazar A., S. Silva A., Silva C., M. Borges C., V. Correia M., S. Santos R. and P. Vilas-Boas J..

W2M2: WIRELESS WEARABLE MODULAR MONITOR - A Multifunctional Monitoring System for Rehabilitation.

DOI: 10.5220/0003785702130218

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2012), pages 213-218

ISBN: 978-989-8425-91-1

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

social care agencies, mainly due to the expected

growing number of elderly people. However, the

impact and interweaving of rehabilitation on the

daily activities of an individual are not sufficiently

emphasized. There exists a need for the investment

of resources aimed to promote quality research in

order to overcome challenges that both therapists

and patients face in real life settings. The

rehabilitation process tends to involve and affect a

number of interpersonal relations that go beyond the

therapist-patient dynamics, shaping communal inter-

dynamics at all levels of social structures; from

families to governments, as illustrated on Figure 1.

Modern technological developments for

rehabilitation must consider such inter-dependency

complexity, thus favoring low-cost adaptable

solutions seeking integration in everyday scenarios

as opposed to rigid strategies meant for limited

usability.

Figure 1: Inter-dynamics of social structures involved in

therapy/rehabilitation process.

An in depth understanding of the human body in

everyday scenarios, advanced skills in physical

assessment, and experience in hands-on

management, allow physiotherapists to manage a

broad range of conditions, with the fundamental goal

of promoting wellness, mobility and independent

function, throughout the lifespan of the individual

(Tate, 2006). Rehabilitation interventions should

specifically address the individual’s impairment; be

sufficiently difficulty to challenge the motor system

and integrate strategies to develop transfer of

performance gains from the training situation to

everyday life (Cirstea and Levin, 2007). Being this

the ideal scenario, an enduring question facing not

only rehabilitation as a field, but healthcare in

general, is whether and to what extent

methodologies and protocols used are effective, and,

if so, whether they are efficient (DeJong, Horn,

Gassaway, Slavin, Dijkers, 2004). Clearly the

answer is dependent on the knowledge of the value

of the outcomes, not only for patients, but also for

end-payers and society as a whole.

A crucial aspect guiding physiotherapist’s

clinical reasoning, and thus design of rehabilitation

intervention, is the assessment of motor

performance; in fact, according to Paten et al. (2010)

implementing therapeutically programs require

accurate clinical and field measurements based on

motor pattern identification not readily available

from traditional tools. Standardized clinical motor

assessments rely on physiotherapists observational

skills, which although may be considered useful,

remains insufficient for reliable measurement of

certain quantitative features (e.g., intersegmental

coordination, quality of movement and smoothness).

Moreover observation-based assessment is subject to

observer induced error, resulting from poor training

(thus mostly confined to more experienced

professionals), personal bias, limited capacity of

human visual perception, just to mention some. The

Rivermean motor assesment (RMA), Fugl-Meyer

motor assessment (FMA), postural assesment scale

for stroke patient (PASS) and the reach performance

scale (RPS) are examples of viable and reliable

measurement instrument commonly used in

physiotherapy. In what refers to evaluation

instrumentation encountered in the laboratory

environment, relevant to the field at hand, one can

refer to EMG, force platforms and complex

image/video analysis systems, that introduce a

degree of objectivity in the interpretation of events,

augmenting the therapist/physician perspectives in

what refers to functional and motor characterization.

However, such resources accessibility in clinical

environment is scarce or null, restricting their

routinely usage from clinical rehabilitation practices.

Physiotherapists clinical practice reality

concerning data gathering and recording is far from

being effective, despite general guidelines

concerning this matter. In fact this important step of

the overall rehabilitation process is many times

absent and when present tends to be mainly

subjective and qualitative, based on the therapist

opinions and patient’s provided information,

regarding movement restoration and overall

progress. Such recordkeeping varies from institution

to institution, from therapist to therapist, and are not

necessarily updated at each session; therefore a

progressive evaluation based on such records

remains subjective to the experience and

interpretation of the reader, which clearly

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

214

contributes to a chain-cycle of deteriorating results,

both in progress and outcomes. Quantitative data

records provide a means for efficient and expedient

analysis of the effectiveness of a therapy on a

patient’s progress, safeguarding from negative

activities that can go unnoticed and unrecorded.

Such approach strengthens and streamlines internal

technological platforms, expanding their coverage

and added-value; promoting the formulation of

standards and protocols for patient progress

monitoring, thus compensating current guidelines.

2 W2M2 ARCHITECTURE

The system hereby presented was originally

intended to address post-stroke upper limb

rehabilitation monitoring. Post-stroke rehabilitation

reality within Portugal (and in most parts of the

world for that matter) is far from being ideal or even

effective, from both the clinically and economical

point of view. In fact, given the growing number of

stroke victims in recent years (portuguese statistical

data reported a mortality rate due to cerebrovascular

diseases per 100.000 inhabitants of 133.9, in 2009

(Portuguese Official Statistics, 2009)) this clinical

condition is considered a major burden on the

healthcare system. Despite the elevated related cost,

there is a general agreement of the importance of

addressing the sequelas of stroke, both due to acute

and sub-acute scenarios, albeit the long-term nature

of the chronic phase. In spite of the well establish

beneficial effects of timely and continuous managed

rehabilitation, there is a clear need to improve its

quantity, quality and overall effectiveness.

Figure 2: Component view of first W2M2 prototype.

At the request of physiotherapists pursuing

alternative therapies for home-based rehabilitation

for post-stroke individuals, a simple inertial

monitoring device was design and implement. The

device was based on commercially available

components that could be assembled in a fast

manner, without extensive knowledge of electronics;

seeking to reduce overdependence on collaborating

engineers. The system concept seen on Figure 2 is

based on a modular approach and almost out-of-box

ready-to-use.

The module’s main component, the Arduino®

FIO, is accessible at low cost and can be

programmed with a reduced learning curve; while its

numerous features such as communication protocols

(IIC, SPI), external interrupts and a number of

analog and digital I/O pins, provide design

flexibility. Additionally, a large amount of open

source resources are available for fast-paced

prototyping and when combined with an XBEE®

based wireless interface, transforms the module in a

portable, wireless adaptable resource. The XBEE®

modules can be readily purchased and are offered

with varying ranges and antennas versions. A

number of inertial and physiological sensors can be

combined with the module, both digital and analog

in nature. Common communication protocols, such

as the IIC and SPI, can be used to chain a number of

modern sensors. The ADXL345 3-axis

accelerometers break-out board was used in this case

for initial testing, permitting data gathering for upper

limb reach based rehabilitation exercises that seek to

generate inertial based biomechanical models for

compensatory movement determination. The sensor

board was encased in a 5.5 x 3 x 2.5 cm standard

project box and fastened on the target individual’s

body through flexible Velcro ended straps (double

sized athletic/rehabilitation tape can also be used for

awkward positions). Figure 3 illustrates a number of

tested sensor positions, utilized for a reach-press-

return post-stroke patient study.

Figure 3: Sensor locations.

The 3-axis accelerometers data was captured at a

frequency of approximately 100 Hz, which was then

buffered and transmitted wirelessly. After package

format verification, the data was processed in

Matlab by applying a simple smoothing strategy and

an auto movement start/end determination based on

a window differentiation function. Supplementary

lines were added for signal stability referencing as

can be observed on Figure 4 a). An example of the

W2M2: WIRELESS WEARABLE MODULAR MONITOR - A Multifunctional Monitoring System for Rehabilitation

215

processed data illustrating the difference between

three independent reach-press.return trials by a

subject without pathology and a subject with

pathology can be seen on Figure 4

Figure

4 b) and c)

respectively.

(a)

(b)

(c)

Figure 4: Sample accelerometry data a) single axis

complete reach-press-return; three independent

accelerometry for reach-press-return b) subject without

pathology c) subject with pathology.

Field usage of the prototype generated a number

of requests by physiotherapist and patients, not to

mention an improved appreciation of the undergone

therapy process. Such gain understanding translated

into a conceptual idea for a modular solution which

allows for an integration of different sensor modules

according to the needs of specific rehabilitation

methodologies. The improved concept considered

multiple modules, either through digital or analog

communication, which could be connected to the

processing unit transforming the original W2M2 into

a multi-purpose monitoring system. Due to the

number of current available sensors, the advantage

to incorporate modules provides a flexibility which

permits the system to adapt to the therapist and

patient on a needs-basis.

At this stage, the W2M2 allows the

interconnection of accelerometer and gyroscope

boards, electrodes, capacitive and galvanic sensors

and also an On/Off sensor, nonetheless the use of

standard communication protocols such as IIC and

SPI permit interconnect to a wide variety of

commercially available sensors which may be added

to address a particular purpose (see Figure 5).

Although, emphasis has been on sensor

interconnection for data gathering, feedback

modules have been considered since the preliminary

stages of the project in order to address home-based

or remote rehabilitation assistance. Digital pre-

processing previous to signal transmission allows for

dataset reduction and optimization for data fusion,

pattern recognition, event detection and feature

extraction strategies; not to mention auto-feedback

response for specific situations as compensatory

movement determination or fall scenarios.

Figure 5: Modular approach of the W2M2.

From a user interface side a number of

alternatives exist for development, such as visual

studio based application, java based application,

scripting languages, even data processing

application such as Excel, Matlab and Labview.

Initial data gathering and visualization was

performed using Processing, an open source

programming language which permits interfacing

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

216

with Arduino® based boards with a minimal

learning curve. The language permits for fast

application implementing and can serve for a

number of specific objectives. In order to provide a

more complete solution, which maintained an

organized database of demographic data and clinical

scores in conjunction with the gathered data, that

said, a VB2010 based application for database

management, data visualization and ease of user

interfacing and control was developed. However,

due to the wireless interfacing through an XBEE®

based USB to serial unit base, software counterparts

can be developed in a wide number of platforms as

mentioned previously. The possibility also exist to

use Bluetooth based XBEE® modules in order to

interface with mobile technology such as cell

phones, or directly with most laptops, tablets and

other devices.

Table 1: Main module and PC interface cost.

Item Cost per Item Total

Arduino FIO 20 20

XBEE modules 30 60

USB – XBEE interface 25 25

Rechargeable Battery 10 10

Misc. (cases, cables, etc) 20 20

Total

- 135

Misc.: miscellaneous

Table 2: Sensor cost examples.

Sensor boards Cost

ADXL 345 3-axis accelerometers 22

IDG500 Dual gyroscopes 34

ITG3200/ADXL345 3-axis acc. & gyr. combo 50

HMC5843 3-axis magnetometer 18

MPR121 12-electrode capacitive touch 9

Acc.: accelerometer; Gyr.: gyroscope

From a cost perspective the W2M2 system is a

low-cost solution, especially when compared to

other physiological and biological signal monitoring

devices available in the market. Devices such as

Biopac, Plux, Biodex and video based solutions

from Qualysis and Viacom are just some examples

of system being used today in rehabilitation and

medical research. These systems retail from

thousands to hundreds of thousands Euros, and tend

to require equally expensive add-ons. In contrast, the

W2M2 can be constructed under 150 Euros and the

sensor components tend to cost less than 50 Euros,

as presented on

Table

1 and

Table

2. In order to

insure good connectivity and measurement

reliability a printed circuit board (PCB) can be

designed for interconnecting the sensor board with

the chosen connector, and sent for manufacturing to

a number of online sites; however such process

tends to cost upwards of 50 Euros for a limited

number of boards. Prototyping boards can offer a

low-cost alternative, especially during initial trials.

Custom sensor solutions, such as electromyography

(EMG) or bioimpedance, can be constructed using

battery based instrumentation amplifiers and

operational amplifiers, and the analog I/O pins of the

main module.

The W2M2 is the key part of a more global data

gathering platform that can be used for home-based

rehabilitation monitoring and as a portable data

acquisition system, as illustrated on Figure 6, which

permits the generation of databases that can be

shared and analyzed for an increase understanding of

human responses to their surroundings under a

number of scenarios (in this case rehabilitation

progress monitoring). There exist a number of

options for remote monitoring interfacing, be it by

periodic data uploading for remote progress analysis,

or real-time one-on-one physiotherapist-patient

sessions. Such interfacing can be achieved through

custom made applications programmed in free

software such as PHP or Visual Basic; or by

utilizing chat applications that allowed for file

sharing and video conferencing.

Figure 6: Complete data acquisition system overview.

3 DISCUSSION

A number of factors have contributed to the fast-

paced development of wearable monitoring

technology of the past decade, where

interdisciplinary collaboration is a factor of

paramount importance. medical/rehabilitation

professionals combining efforts with

scientist/engineers has become commonplace,

W2M2: WIRELESS WEARABLE MODULAR MONITOR - A Multifunctional Monitoring System for Rehabilitation

217

advancing areas which traditionally were slow to

adapt. Some systems are now appearing capable of

human movement analyze and for physiotherapy

assistance (Zhou and Hu, 2007; Pérez et al., 2010);

however, their institutional oriented approach makes

them impractical for widespread usage due to highly

technical learning curves and/or required

accommodations size and electrical requirements;

not to mention prohibitive costs. The system

presented seeks to ease the data gathering process by

offering an adaptable low cost alternative that can be

used and modified by physician, therapist and even

patients (for home-based and/or remote solutions).

Its portability allows for data gathering in a number

of diverse scenarios including home-based

rehabilitation, or even daily-activity, monitoring;

altering the therapist-patient dynamics by extending

the rehabilitation process.

Currently, visually based biomechanical models

seem to dominate the field, probably due to

traditional appreciation methodologies for patient

progression; however, inertial, chemical, electrical

sensors are broadening the perceptual capacities of

current rehabilitation practices, introducing the need

for new approaches and models. Through data

gathering practices by physicians, therapist and even

patients, a body of comparable datasets can be

generated for the formulation of statistical and

analytical methods that can reveal quantifiable

methods that can contribute in the diagnosis,

treatment and follow-up of numerous conditions.

4 CONCLUSIONS

A data acquisition system was designed and

developed as a low-cost, fast implementable

alternative for rehabilitation monitoring. Although

the device was originally thought for post-stroke

upper limb rehabilitation monitoring, its flexibility

and adaptability allowed usage in a number of

monitoring objectives. The device seeks to ease the

data gathering process by therapist and others, in

order to facilitate the development of quantifiable

methodologies and protocols.

ACKNOWLEDGEMENTS

The authors would like to thank the Foundation for

Science and Technology of Portugal for their

support of some of the PhD students involved in this

article (SFRH/BD/61396/2009 and SFRH/BD/609

29/2009). Additionally, the authors would like to

acknowledge the contribution of all volunteers that

took part of the testing procedures.

REFERENCES

Cirstea, M.; Levin, M. 2007. Improvement of arm

movement patterns and end-point control depends on

type of feedback in stroke survivors. Neurorehabil

Neural Repair. 21: 398-411.

DeJong, G., Horn, S., Gassaway, J., Slavin, M., Dijkers,

M. 2004. Toward a taxonomy of rehabilitation

interventions: using an inductive approach to examine

the “black box” of rehabilitation. Arch Phys med

Rehabil. 85: 678-86.

EMB Technical Committee on Wearable Biomedical

Sensors & Syst. Available http://tc-wearable-sensors.

embs.org/

Laerhoven, K. V., Lo, B., Ng, J., Thiemjarus, S., King, R.,

Kwan, S., Gellersen, H. W., Sloman, M., Wells, O.,

Needham, P., Peters, N., Darzi, A., Toumazou, C.,

Yang, G.-Z. 2004. Medical Healthcare Monitoring

with Wearable and Implantable Sensors. Paper

presented at the 3rd Int. Workshop on Ubiquitous

Computing for Healthcare App.

Paten, S., Hughes, R., Hester, T., Stein, J., Akay, M., Dy,

J., Bonato, P.: Tracking Motor Recovery in Stroke

Survivors Undergoing Rehabilitation Using Wearable

Technology. 32nd Annual International Conference of

the IEEE EMBS, pp. 6858 – 6861. IEEE Press, (2010)

Pérez, R., Costa, Ú., Torrent, M., Solana, J., Opisso, E.,

Cáceres, C., Tormos, J., Medina, J., Gómez, E. J.

2010. Upper Limb Portable Motion Analysis System

Based on Inertial Technology for Neurorehabilitation

Purposes. Sensors, 10.

Portuguese Official Statistics. Health Database on Rate of

potential years of life lost due to cerebrovascular

diseases per 100 000 inhabitants, 2009.

Tate, D. 2006. The state of rehabilitation research:art or

science?. Arch Phys Med Rehabil. 87: 160-6

Tufail, F., and Islam, M. H. 2009. Wearable Wireless

Body Area Networks. Paper presented at the

International Conference on Information Management

and Engineering.

Zhou, H., Hu, H. 2007. Upper limb motion estimation

from inertial measurements. International Journal of

Information Technology, 13(1).

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

218