Zoom out Search Issue

ManualsBrandsContents Manualsaudio & home theatreZoom in

John Edwards

[

special

REPORTS

]

Digital Object Identifier 10.1109/MSP.2014.2373231

Date of publication: 12 February 2015

Signal Processing Drives a Medical Sensor Revolution

ensor technology’s impact on

health care is growing rapidly.

New applications are appearing

almost daily. Wireless sensors

are now used in an ever-grow-

ing number ways, such as monitoring glu-

cose levels in diabetics, recording and

tracking heart irregularities, and diagnos-

ing infectious diseases. Linking sensors to

mobile phones has made wearable sensors

a reality, allowing individuals to monitor

not only chronic diseases but also their

lifestyle activities.

“There’s a lot happening in health

monitoring,” says Andreas Spanias, a pro-

fessor in the Arizona State University

School of Electrical, Computer, and Ener-

gy Engineering. “Integrated sensors, on

mobile phones, for example, can monitor

vital signs, such as heart rates, breathing

activity, oxygenation, and blood pressure,”

says Spanias, who is also the founder and

director of the university’s Sensor Signal

and Information Processing (SenSIP) Cen-

ter, a National Science Foundation (NSF)

Industry and University Cooperative Re-

search Center (NSF/UCRC).

Spanias says that signal processing is

essential to optimal sensor operation and

performance. “Our industry collaborators

build inexpensive sensors, and signal pro-

cessing improves precision and event de-

tection using machine learning and

fusion,” he notes. “Even if the data is noisy

or contains artifacts, signal processing can

reduce noise effects. Signal processing al-

gorithms, for example, will make wireless

health monitoring more accurate and reli-

able. Signal processing makes it possible to

use data from several sensors and combine

the information appropriately to maximize

the probability of correct detection.”

Signal processing algorithms are

likely to become even more essential to

wireless health-care sensor development

in the years to come. The technology is

now entering a new phase made possible

by the development of microscopic

nanosensors and nanorobots designed

for insertion into bodily tissues and the

bloodstream. “With so many sensors in

the body, and the large volumes of data

they will be transmitting, how do you

fish out the information that you need?”

Spanias asks. “Signal processing and bio-

medical informatics will have a big role

in that area, and algorithms will enable

reliable prediction of disease and incen-

tivize healthy lifestyles.”

INTRABODY NETWORKS

A system of wirelessly networked intra-

body sensors and actuators could lead to

revolutionary new applications in health-

care monitoring, potentially creating in-

novative approaches to the treatment of

an almost endless number of diseases,

both major and minor. Yet an important

obstacle to the development of reliable

intrabody sensor/actuator networks is the

fact that most health-care sensor net-

work research to date has focused on

communication along the body surface

via devices linked through traditional

electromagnetic radio-frequency (RF)

transmissions. Such technology has sig-

nificant limitations for intrabody system

developers, however, due to the physical

nature of propagation within the human

body, which is composed primarily of wa-

ter, a medium through which RF electro-

magnetic waves do not easily propagate.

Researchers at Northeastern Univer-

sity in Boston, in collaboration with re-

searchers at the University of Catania

and the Sapienza University of Rome, are

hoping that by taking a novel approach

to wireless sensor communication—ul-

trasonic networking technology—they

can make intrabody sensors and actua-

tors an accurate and reliable technology.

The researchers are currently pursuing

a closed-loop combination of mathemat-

ical modeling, simulation, and experi-

mental evaluation to determine the

practicality of using ultrasonic network-

ing in human tissues. “A major chal-

lenge is creating a waveform that’s

resistant to the effects of multipath and

scattering,” says team member Tommaso

Melodia, an associate professor in North-

eastern University’s Department of Elec-

trical and Computer Engineering.

The magnitude and direction of a re-

flected wave depends on the orientation

of the boundary surface as well as on the

acoustic impedance of the tissue. Scat-

tered reflections happen whenever an

acoustic wave encounters an object that’s

relatively small in relation to its wave-

length or meets a tissue with an irregular

surface. “At the receiver, basically, you re-

ceive a combination of multiple replicas

of the same signal,” Melodia says. “You

have to create a receiver that can differ-

entiate between these various signals; it

basically needs to be able to record the

original signal from multiple replicas

that it’s receiving.”

LINKING SENSORS TO

MOBILE PHONES HAS

MADE WEARABLE SENSORS

A REALITY, ALLOWING

INDIVIDUALS TO MONITOR

NOT ONLY CHRONIC

DISEASES, BUT ALSO THEIR

LIFESTYLE ACTIVITIES.

THE WORLD’S NEWSSTAND

THE WORLD’S NEWSSTAND