17-03-2014, 10:14 PM
Abstract—Wireless sensor nodes (WSNs) have recently evolved
to include a fair amount of computational power, so that
advanced signal processing algorithms can now be embedded
even in these extremely low-power platforms. An increasingly
successful field of application of WSNs is tele-healthcare, which
enables continuous monitoring of subjects, even outside a medical
environment. In particular, the design of solutions for automated
and remote electrocardiogram (ECG) analysis has attracted
considerable research interest in recent years, and different
algorithms for delineation of normal and pathological heart
rhythms have been proposed. In this paper, some of the most
promising techniques for filtering and delineation of ECG signals
are explored and comparatively evaluated, describing their implementation
on the state-of-the-art IcyHeart WSN. The goal of this
paper is to explore the trade-offs implied in the different settings
and the impact of design choices for implementing “smart” WSNs
dedicated to monitoring ECG bio-signals.
Index Terms—Embedded Systems, Bio-medical Signal Processing,
Wireless Sensor Networks, ECG Delineation, ECG Filtering.
I. INTRODUCTION
The aging of world population, coupled with others factors
such as the predominance of unhealthy lifestyles, is enlarging
the impact of cardiovascular disorders, which now represent
the most common global cause of death [1]. Supervising patients
affected by cardiopathy is putting an increasing pressure
over health care systems.
Electrocardiograms (ECGs), which measure the electrical
activity of the heart, are the primary instruments for monitoring
the heart activity and for early detection of heart
pathologies. A breakthrough in the practice of ECGs recording
and analysis has been possible thanks to Wireless ECG Sensor
Nodes (WESNs) [2][3]: miniaturized, wearable ECG devices
that are able to wirelessly transmit relevant data, allowing the
continuous and autonomous monitoring of subjects.
The continuous progress in semiconductor technology has
enabled the emergence of “smart” WESNs that, in addition to
acquisition and transmission, perform an automated diagnosis
by interpreting ECG data [4][5]. Smart WESNs applications
The research illustrated in this paper has been supported by the IcyHeart
European project (FP7, capacities. Proj. num. 286130.)
usually implement algorithms to filter acquired signals, delineate
individual heart beats to retrieve their characteristics
and classify them to detect pathologies. The first two steps
(filtering and delineation), focus of this paper, have the most
challenging real-time constraints, because they deal with the
manipulation and analysis of digital signals [6][7][8], as opposed
to parameters of whole heart beats.
While proposals are described in the literature to implement
embedded delineation and filtering [9][10], a comparative evaluation
of different methods is still lacking, and this paper aims
at filling this gap. In this work we specifically target WESN
solutions, only considering methods that can be implemented
on these resource-constrained platforms, which usually only
support integer arithmetic and present a small memory size.
While delineation could be conceptually performed on raw
signals, we do not consider this option in this work, as realworld
acquisitions are always corrupted by both low-frequency
noise sources (caused by subjects’ respiration and perspiration)
and high-frequency ones (due to muscular contractions or
misplacements of electrodes), mandating a filtering phase [6].
to include a fair amount of computational power, so that
advanced signal processing algorithms can now be embedded
even in these extremely low-power platforms. An increasingly
successful field of application of WSNs is tele-healthcare, which
enables continuous monitoring of subjects, even outside a medical
environment. In particular, the design of solutions for automated
and remote electrocardiogram (ECG) analysis has attracted
considerable research interest in recent years, and different
algorithms for delineation of normal and pathological heart
rhythms have been proposed. In this paper, some of the most
promising techniques for filtering and delineation of ECG signals
are explored and comparatively evaluated, describing their implementation
on the state-of-the-art IcyHeart WSN. The goal of this
paper is to explore the trade-offs implied in the different settings
and the impact of design choices for implementing “smart” WSNs
dedicated to monitoring ECG bio-signals.
Index Terms—Embedded Systems, Bio-medical Signal Processing,
Wireless Sensor Networks, ECG Delineation, ECG Filtering.
I. INTRODUCTION
The aging of world population, coupled with others factors
such as the predominance of unhealthy lifestyles, is enlarging
the impact of cardiovascular disorders, which now represent
the most common global cause of death [1]. Supervising patients
affected by cardiopathy is putting an increasing pressure
over health care systems.
Electrocardiograms (ECGs), which measure the electrical
activity of the heart, are the primary instruments for monitoring
the heart activity and for early detection of heart
pathologies. A breakthrough in the practice of ECGs recording
and analysis has been possible thanks to Wireless ECG Sensor
Nodes (WESNs) [2][3]: miniaturized, wearable ECG devices
that are able to wirelessly transmit relevant data, allowing the
continuous and autonomous monitoring of subjects.
The continuous progress in semiconductor technology has
enabled the emergence of “smart” WESNs that, in addition to
acquisition and transmission, perform an automated diagnosis
by interpreting ECG data [4][5]. Smart WESNs applications
The research illustrated in this paper has been supported by the IcyHeart
European project (FP7, capacities. Proj. num. 286130.)
usually implement algorithms to filter acquired signals, delineate
individual heart beats to retrieve their characteristics
and classify them to detect pathologies. The first two steps
(filtering and delineation), focus of this paper, have the most
challenging real-time constraints, because they deal with the
manipulation and analysis of digital signals [6][7][8], as opposed
to parameters of whole heart beats.
While proposals are described in the literature to implement
embedded delineation and filtering [9][10], a comparative evaluation
of different methods is still lacking, and this paper aims
at filling this gap. In this work we specifically target WESN
solutions, only considering methods that can be implemented
on these resource-constrained platforms, which usually only
support integer arithmetic and present a small memory size.
While delineation could be conceptually performed on raw
signals, we do not consider this option in this work, as realworld
acquisitions are always corrupted by both low-frequency
noise sources (caused by subjects’ respiration and perspiration)
and high-frequency ones (due to muscular contractions or
misplacements of electrodes), mandating a filtering phase [6].