See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/210290210

Neonatal Monitoring Technologies: Design for Integrated Solutions, Book

chapter: Neonatal Infrared Thermography Imaging

Chapter · January 2011

CITATIONS READS

4 13,770

5 authors, including:

Thorsten Orlikowsky Steffen Leonhardt

University Hospital RWTH Aachen RWTH Aachen University
217 PUBLICATIONS   1,638 CITATIONS    646 PUBLICATIONS   5,495 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Simplified Engineering Approaches (SEA) Series View project

Focal and non-Focal Epilepsy Localisation View project

All content following this page was uploaded by Abbas K. Abbas on 19 May 2014.

The user has requested enhancement of the downloaded file.

Neonatal Monitoring
Technologies:
Design for Integrated Solutions
Wei Chen
Eindhoven University of Technology, The Netherlands

Sidarto Bambang Oetomo

Máxima Medical Center, The Netherlands, and Eindhoven University of
Technology, The Netherlands

Loe Feijs
Eindhoven University of Technology, The Netherlands
Managing Director: Lindsay Johnston
Senior Editorial Director: Heather A. Probst
Book Production Manager: Sean Woznicki
Development Manager: Joel Gamon
Development Editor: Mike Killian
Acquisitions Editor: Erika Gallagher
Typesetter: Russell A. Spangler
Cover Design: Nick Newcomer, Lisandro Gonzalez

Published in the United States of America by

Medical Information Science Reference (an imprint of IGI Global)
701 E. Chocolate Avenue
Hershey PA 17033
Tel: 717-533-8845
Fax: 717-533-8661
E-mail: cust@igi-global.com
Web site: http://www.igi-global.com

Copyright © 2012 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in
any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher.
Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or
companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark.

Library of Congress Cataloging-in-Publication Data

Neonatal monitoring technologies: design for integrated solutions / Wei Chen, Sidarto Bambang Oetomo, and Loe Feijs,
editors.
p. cm.
Summary: “This book presents a unique integration of knowledge from multidisciplinary fields of engineering, industrial
design, and medical science for the healthcare of a specific user group”-- Provided by publisher.
Includes bibliographical references and index.
ISBN 978-1-4666-0975-4 (hardcover) -- ISBN 978-1-4666-0976-1 (ebook) -- ISBN 978-1-4666-0977-8 (print & perpetual
access) 1. Neonatology--Technological innovations. 2. Newborn infants--Medical examinations. 3. Premature infants--
Hospital care. I. Chen, Wei, 1978 Sept. 30- II. Oetomo, Sidarto Bambang, 1951- III. Feijs, L. M. G. (Loe M. G.)
RJ251.N37 2012
618.92’010284--dc23
2011048667

British Cataloguing in Publication Data

A Cataloguing in Publication record for this book is available from the British Library.

All work contributed to this book is new, previously-unpublished material. The views expressed in this book are those of the
authors, but not necessarily of the publisher.
84

Chapter 5
Neonatal Infrared
Thermography Monitoring
Abbas K. Abbas Katrin Jergus
RWTH Aachen University, Germany University Children´s Hospital and RWTH
Aachen University, Germany
Konrad Heiman
University Children´s Hospital and RWTH Thorsten Orlikowsky
Aachen University, Germany University Children´s Hospital and RWTH
Aachen University, Germany
Steffen Leonhardt
RWTH Aachen University, Germany

ABSTRACT
For critically ill preterm infants, there is a clinical need for contact-free monitoring technologies, which
would eliminate discomfort and potential harm (e.g., necrosis) due to adhesive electrodes, temperature
and saturation sensors. Hence, this chapter focuses on non-contact physiological monitoring of infants
based on infrared (IR) thermography. This technique has the potential to replace the conventional
temperature sensing by detecting radiated thermal energy emitted from the baby’s surface according
to black-body radiation principle. This allows the application of a less invasive method giving more
detailed information about the thermoregulation status of newborn infants. As an illustrative example,
an investigation into thermoregulation physiology during kangaroo care method has been chosen to
illustrate the benefit of this method for standardized neonatal intensive care unit (NICU) procedures.
Furthermore, this technique may have a large impact on non-contact respiratory monitoring, as it al-
lows quantitative evaluation of the heat transfer processes over nostrils region. Moreover, the ability to
detect infrared respiration (IRTR) signature with thermography imaging, will pave the road toward a
non-contact breathing monitoring. This in turn will influence the development efforts for wireless and
smart incubator solutions.

DOI: 10.4018/978-1-4666-0975-4.ch005

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
Neonatal Infrared Thermography Monitoring
1. INTRODUCTION
Despite the incapacity of the preterm neonate to
adequately control body core temperature and to
compensate large temperature gradients, he/she
can adapt to minor changes in the environmental
temperature if they are incubated inside a thermo-
regulated micro-climatic zone. Hence, one major
point in preterm infant care is to avoid cold stress
and to keep the neonate’s body within the still
small thermoneutral zone without affecting other
physiological functions.
Clinically, neonates require a typical skin
temperature within the range of 35.5 °C … 37.2
°C and a core temperature of approximately 37
°C (Asakura, 2004). Therefore, the maintenance
of this temperature zone is a highly differentiated
process, involving lipolysis and gluconeogenesis.
This is associated with consumption of energy,
oxygen and glucose. The more energy is needed
for maintaining a constant body temperature,
the less energy is available for other proceed-
ings such as growth, brain development, or lung
maturation. Basically, hypothermia causes high
oxygen consumption with a left shift of the oxy-
gen dissociation curve, resulting in acidosis, less
oxygen supply in tissues, and vasoconstriction
(Bissinger & Annibale, 2010). For neonates, es-
pecially if they are preterm, this implicates two
consequences: The urgent need for analyzing
body temperature constantly, and the provision
of a thermoneutral zone, accomplished by an
incubator with a proper temperature and humid-
ity. Basically, the thermoneutral zone is defined
as the area of surrounding temperature in which
lowest energy exchanges and therewith minimal
oxygen consumption takes place.
Clinically, the standards to measure central
and peripheral temperature are by cable-bound
temperature sensors which are placed either into
the rectum or are attached to the skin of the infant,
respectively. However, as adhesive connections
cause mechanical stress to the very sensitive skin
of the infant and lying on cables may ultimately
cause ulcers, there is a growing demand to install
cable-free monitoring techniques into incubators
or generally within the intensive care stations.
Such a contact-free monitoring modality is
infrared (IR) imaging. The first documented ap-
plication of infrared imaging in medicine was
in 1956 (Bissinger & Annibale, 2010; Flenday,
2003) when female breast cancer patients were
examined for asymmetric hot spots and vascularity
in IR images of their breasts. Since then, several
research findings have been published and the
1960s witnessed the first surge of medical ap-
plication of the IR technology with breast cancer
detection as the primary practice (Flenday, 2003;
Bohnhorst, 2010). However, IR imaging has not
been widely recognized in medicine yet largely
due to high costs and the premature use of the
technology. In neonatal care, Clark and Stothers
(Clark, 1980) were the first to work on mapping
temperature distribution of neonatal skin surface,
which can be considered as the first trial of neo-
natal infrared thermography imaging.
An interesting study investigating non-invasive
temperature monitoring based on IR imaging in
neonates was performed by Kimberly and Horns
(2003). In this work, skin surface temperature
gradients using infrared thermography were
recorded. In fact, the authors investigated the
instability of neonatal temperature as main index
for morbidity in Very Low Birth Weight (VLBW)
infants using IR thermography for detecting cold
stress and other related physiological parameters
as changes in peripheral perfusion. Possibly, this
pilot project will eventually provide a descrip-
tion of existing patterns for thermoregulatory
control in the first day of life, substantiate for
non-invasive technologies, and explores the feasi-
bility of infrared thermographs in infants who are
care for in a special environment (Horns, 2003;
Clark, 1980). Hence, these findings may serve as
data for future research in predicting early signs
of infection and impeding shock before serious
blood pressure changes are demonstrated on the
bed-side monitors.
85

Figure 1. High-resolution IR snap shot of an infant (a), with temperatures in a pseudocolor represen-
tation. Note the rather cool background inside the incubator,(b) schematic drawing of the early trail
to register new-born infant temperature mapping with IR thermography, as performed by Clark and
Stothers (1980) (© MedIT, 2008).
Figure 1(a) shows a pseudocolor thermogram
snap-shot of a preterm infant recently taken at
RWTH Aachen University Hospital. In addition,
Figure 1(b) illustrates the basic measurement
setup, as proposed by Clark and Stothers in
1980 (Clark, 1980; Cannon, 2004). As already
mentioned, the main temperature distribution
profiles on the infant body, as achieved by Horns
and again in a later trial by Adams et al. (Adams,
2000), may be used to calculate mean heat loss
in postnatal life through the evaporative route.
At that time, these methods were still technically
challenging due to low thermal resolution of the
IR detectors and a dedicated method to obtain a
precise measurement of neonate surface geometry
(Horns, 2003; Cannon, 2004; Adams, 2000).
2. SOME BACKGROUND ON
NEONATAL INCUBATORS
In the following, the authors will provide some
background information on history, thermodynam-
ics (physics) and requirements for intensive care
of neonates. This background is to give technical
and clinical overview to be related with applica-
86
Neonatal Infrared Thermography Monitoring
tion of the infrared thermography imaging within
this medical environment (closed convective
incubator).
2.1. A Brief History of
Neonatal Incubators
Protecting preterm infants against excessive heat
loss improves their chances for survival, reduces
their bodies’ need to produce heat by metabolic
work, and eliminates the problems associated with
rewarming of hypothermic infants. Note that the
infant incubation-prospect may be traced back in
time to ancient Mesopotamia (2000 B.C), where
wet nursing with protected thick-blanket was
used (as many viewing of clay-tablets (McIntosh,
2005) (Figure 2).
Incubators used for human infant care did not
exist until the late 1870s. In fact, the incubator
was part of the primitive technology in Professor
Tarnier’s method for enhancing the prematurely
born babies’ survival rate (Aylott, 2006; Cramer,
2005; Silverman, 2007). The historical background
of incubators is the history of neonatal medicine,
which includes the bizarre but critically important
use of incubators to display human infants in
Neonatal Infrared Thermography Monitoring

Figure 2. Mesopotamian clay tablet (a) depicting a woman with a child, third or second millennium
B.C.E., (b) Schematic of egg-incubation method used by ancient Egyptian to store eggs in good thermal
environment, in which this ancient method inspired to develop infant incubator model (Zev Radovan/
Land of the Bible Picture Archiv (McIntosh, 2005)

sideshow exhibits. Other sources provide excellent
reviews of this period (McIntosh, 2005).
Throughout intrauterine life, the heat-energy
production by the fetus results in a fetal tem-
perature that is about 0.5 - 0.7 °C higher than the
maternal milieu temperature (~ 37.2 °C -37.5 °C).
Shortly after birth, the newborn infant is exposed
to air and surfaces, which have a much lower
temperature than that previously experienced
inside his mother’s uterus. Eventually, the skin
surface at that time is covered with amniotic
fluid, causing heat loss through evaporation in
an environment with a low water-vapor pressure.
Accordingly, the infant’s body temperature is
decreased, and the rate of this reduction is influ-
enced by the temperature of the environmental
air in the delivery room and the velocity of its
flow. This may induce thermogenic responses that
increase basal metabolic (BMR) heat production,
and the skin blood circulation may diminish to
lower the heat losses (Adams, 2000; Aylott, 2006;
Cramer, 2005).
Basically, thermal balance in newborn infants
depends on the heat transfer regulation between
the infant and the surrounding environment. This
physical energy transfer is directly related to the
temperature and humidity of the surrounding air
(the water vapor pressure), the flow velocity of
that air, the temperatures of the surfaces facing
the infant (e.g., hood ceiling, walls of room or
incubator, and bedding material (mattress)), and
the surface temperature in contact with the infant
(Silverman, 2007; Chadron, 2006).
Following delivery, the immediate interven-
tions needed to avoid body cooling are to wipe
the amniotic fluid from the skin surface to lower
the loss of heat through evaporation and to cover
the infant with a warm and dry towel or blanket,
or both, to reduce the exposure of the infant’s
skin to the environment. The infant born at term
or moderately preterm can be covered with a
blanket and then placed on the mother’s chest,
but extremely preterm infants usually need other
measures to maintain their body temperature - usu-

87

Figure 3. Schematic of “von Ruehl” warming tub incubator design (1835), (a) lateral cross section view,
and (b)frontal cross-section, according to Silverman (2007) (© MedIT, 2010)
ally placement in an incubator or under a radiant
heater, and, if necessary, mechanical ventilation
with warm and humidified gas (Silverman, 2007;
Chadron, 2006; Knobel, 2010).
Prior to the nineteenth century, keeping small
babies warm meant swaddling with multiple layers
of cloth, providing body contact with the mother,
or placement of the infant near a warm, roaring
fireplace. Such classic thermal care served lusty,
healthy babies well, but was inadequate to provide
for the special needs of premature or otherwise
enfeebled newborns (Blackburn, 2001). These
special needs were not met because, until the last
century, there was almost no recognizable major
medical or social commitment toward enhanc-
ing the survival of babies born prematurely. The
first attempt to improve the thermal protection of
newborns was reflected in a warming tub model
developed in 1835 by Johann Georg von Ruehl
(1769–1846) in Russia (Museux, 2008) (Figure
3). Basically, the “von Ruehl” tub was simply a
double-walled sheet-iron open cradle that was
kept warm by filling the space between the walls
with warm water. Variations on “von Ruehl’s”
design were subsequently developed throughout
Europe, and this type of primitive open incubator
remained a standard device for care until 1878.
88
Neonatal Infrared Thermography Monitoring
Additionally, Prof. Tarnier developed another
incubator model based on von Ruehl design, in
which it increases thermal protection to the new-
born infant inside incubator environment more
than von Ruehl model (Figure 4).
The incubator modernization process has not
only led to marked improvement in the thermal
protection provided infants, but it also has been
pivotal in increasing our knowledge of diseases
distinctive to newborn babies. Consecutively,
each increment of knowledge has required
manufacturers to modify incubators in order to
provide solutions to problems the new scientific
discovery imposed (Blackburn, 2001; Cramer,
2005).
For example, when electric fans became
available and forced air convection systems were
developed, incubator heating became so improved
that for the first time infants could be undressed
during care. This fact, along with the use of new
transparent plastics in the construction of incubator
walls, allowed neonatologists to make observa-
tions that led to in-depth description of illnesses
about which little was known when infants were
hidden within the predominately opaque wooden
and metal chambers of the past (Hyman, 2006;
Jamsa, 1998).
Neonatal Infrared Thermography Monitoring
Figure 4. Concept design of the Tarnier incubator model (1880), according to Silverman (2007) (©
MedIT, 2010)
However, while applying of clear plastic or
Plexiglas® in incubator construction enhanced the
ability to observe infants, its poor insulating quali-
ties made the task of maintaining the incubator
chamber in a warm and stable state more difficult.
Thus far, the visibility improvements led to new
lifesaving therapies, such as the intravenous fluids
administration, ventilator using, and the develop-
ment of new diagnostic procedures and surgical
interventions. Additionally, the transparent plastic
walls that provided caretakers with visual access
to sick babies during these therapeutic processes
also served as barrier (Silverman, 2007; Museux,
2008; Libert, 1997).
2.2. Heat Transfer Mechanisms
in Preterm Infants
In principle, the heat exchange between the infant
and its environment occurs through the skin and
through the respiratory tract. The introduction of
new techniques to measure the rate of evaporation
from the skin and from the respiratory tract has
permitted the heat energy loss through evaporation
to be quantified. It is also possible to calculate the
heat loss through other modes of heat exchange,
such as radiation, convection, and conduction. In
addition, to determine the overall heat loss per unit
surface area from the total neonate’s body surface
area, it is also necessary to take into consideration
the proportions of surface area participating in
the different modes of heat exchange (Figure 5)
(Chadron, 2006; Knobel, 2010).
Additionally, the heat exchange modes between
the infant’s respiratory tract and the environment
can be determined. Essentially, the heat exchange
through evaporation, radiation, convection, and
conduction can be calculated with knowledge of
the transepidermal water loss (TEWL), the tem-
perature of the infant’s skin (Tskin), the temperature
of the walls facing the infant (Twall), the tempera-
ture of the ambient air (Tamb), the temperature of
the material on which the infant is placed (Tmatress),
and characteristics of the material in the infant’s
89

Figure 5. Heat transfer mechanisms of a newborn inside an open radiant warmer (© MedIT, 2010)
environment. Principally, all these parameters will
be used in the following sections (Museux, 2008;
Libert, 1997).
In case of absence of forced convection, and if
the effect of thermal diffusion can be disregarded,
the process of water exchange through a station-
ary water-permeable surface can be expressed in
terms of the vapor-pressure gradient immediately
adjacent to the surface. The evaporation rate (ER,
in g/m2h) from the infant’s skin can thus be de-
termined by a method based on calculation of the
water vapor pressure in the layer of air immediately
adjacent to the neonate’s skin surface. Note that
the relationship between the vapor pressure and the
distance from the evaporating surface is linear in
this region. If the gradient in this layer is known,
the amount of water evaporated per unit time can
be calculated (Museux, 2008; Rojas, 2006).
Using so called “Evaporimeter” (e.g., Ser-
voMed® AB, Varberg, Sweden) allows to quickly
measuring free evaporation without disturbing the
infant. Transepidermal water loss, which is a mean
value of cutaneous water loss, can be calculated
according to
90
Neonatal Infrared Thermography Monitoring
TEWL = 0.92 x ER + 1.37 [g/m2h] (1)
where ER is the arithmetic mean of the ER mea-
sured from the chest, an interscapular area, and
buttock region. Therefore, the heat exchange
through evaporation (Hevap) can be evaluated if
TEWL is known, according to the equation
Heavp = K1 x TEWL (3.6x103)-1 [W/m2] (2)
where k1 is the latent heat of evaporation (2.4
× 103 J/g), W is Watt and 3.6 × 103 is the cor-
rection factor for time (seconds). Figure 6 il-
lustrates the relationship of the transepidermal
water loss, the gestational age (GSA) and the
post-natal age.
Basically, the heat exchange through radiation
(Hrad) can be determined if the mean temperature
of the skin (T1, in °K (Kelvin)) and the mean
temperature of the surrounding walls (T2, in K)
are known:
Hrad = S0 . e1. e2.(T41- T42) [W/m2] (3)
Neonatal Infrared Thermography Monitoring
Figure 6. Relationship between the transepidermal water-loss, the gestational age and the postnatal age
for preterm infants (modified from Cloherty, 2008)
where S0 is the Stefan-Boltzmann constant (5.7 ×
10-8 W/m 2K4), e1 is the emissivity of the skin, and
e2 is the emissivity of the surrounding walls (0.97).
Heat exchange through convection (Hconv)
can be calculated if the mean temperature of the
skin (T1, in K) and the mean temperature of the
ambient air (T3, in K) are known:
H conv = k2 (T1-T3) [W/m2] (4)
where k2 is the convection coefficient (2.713
W/m2 K). This coefficient for convection has
been used in many of the studies referred in this
chapter, but it has usually been determined in
measurements on adult human skin. A convec-
tion coefficient suggested as being more valid
for newborn infants can alternatively be used,
and Hconv is then 48% higher.
Heat exchange through conduction (Hcond)
can be determined if the temperature of the skin
(Tskin; K) and the temperature of the bed (Tbed °C)
are known:
Hconv= k0 (Tskin-Tmatress) [W/m2] (5)
In this equation, k0 is a conductive heat transfer
coefficient. With the thermal conductivity char-
acteristics of most regular mattresses, the heat
91

loss through conduction is very low (Chadron,
2006). Hence, the extent of overall heat exchange
between the body surface area and the environment
depends on the magnitude of the different heat
exchange routes, on the position and geometry of
the body, and on the magnitude and frequency of
body movements. Because different modes of heat
exchange are unequally influenced by changes
in body position, the relative contribution of dif-
ferent modes of heat exchange might vary with
time (Silverman, 2007; Chadron, 2006; Cloherty,
2008). Heat exchange is often described per unit
area of body surface exposed to the ambient air
or facing the walls of the incubator.
2.3. Requirements for an Optimal
Thermal Environment
In neonatal intensive care units (NICU), babies
are often nursed naked for easier handling and ob-
servation. This would normally result in increased
heat loss, as ambient conditions are usually too cold
for a newborn to maintain heat balance. Therefore,
these babies have to be protected by a sufficiently
warm environment. To avoid any heat loss from
the baby, the immediate environment referred as
the “micro-climate” must be controlled in such a
way that the baby’s own heat production and all
heat losses are in balance. The “micro-climate”
can be controlled with special warming therapy
equipment such as incubators, open radiant warm-
ers, heated cots etc (Ginaliski, 2007).
In fact, infant incubator forms a closed thermal
system (“micro-climate”) in which temperature
and humidity are feedback-controlled. The incu-
bator’s closed design allows a higher humidity
and oxygen concentration than the ambiance.
Individual incubator designs should comply with
the International Standards for Incubators. In the
incubator, the maximum heat transfer to the baby
is normally not more than 100 W/m2 as opposed
to radiant warmers which can provide up to 600
W/m2. This factor explains the limited ability of
an incubator to warm up a hypothermic baby.
92
Neonatal Infrared Thermography Monitoring
One advantage of the incubator is that it provides
protection against bacterial infection due to the
design principle of a closed hood and altered ambi-
ent air. On the other hand, the incubator creates a
physical barrier between the baby and the nurse or
parents (Cloherty, 2008, Ginaliski, 2007, 2008).
Fundamentally, convection is the main mode
of heat transfer used by incubators to transfer heat
to the baby. When the incubator air temperature is
cooler than the baby’s skin temperature, heat loss
from the baby is achieved by the high velocity of
air flow. Therefore, modern incubators should have
a minimal air flow velocity (6 to 8 cubic decimeter
per second (cds)) (Chadron, 2006; Rojas, 2006).
However, a low air flow velocity reduces the
ability to effectively warm up hypothermic babies
within the shortest possible time.
Figure 7(a) illustrates the convection type of
infant incubator based on the double wall ap-
proach: this means that the incubator hood will be
made of dual layer plexiglass material to minimize
the heat loss through convection and radiation.
Moreover, Figure 7(b) displays the convective air
stream from the incubator actuators (heater and
fan) to the enclosed internal geometry of incubator.
Basically, the air temperature distribution varies
throughout the incubator geometry, due to the con-
vective and radiative heat loss mechanisms during
air circulation. In incubators, the air temperature
is controlled by a thermostat. Some incubators
allow dual control modes (air-temperature and
skin-temperature-control).
The incubator air temperature can be set by
the user within the normal range of 28 to 39 °C.
To prevent unacceptably high temperatures due
to malfunction of any part, typically an indepen-
dent second safety thermostat is installed (Rojas,
2006; Cloherty, 2008). The operator can choose
manual control of the incubator air temperature
or the baby’s skin temperature control (Servo
Control Mode). The philosophy behind controlling
air temperature is to provide a constant environ-
mental temperature in which the baby’s heat
production and heat losses caused by environ-
Neonatal Infrared Thermography Monitoring
Figure 7. Schematic drawing of a typical convective incubator (a) currently used in many NICU (b)
Cross-section of infant incubator showing the convective heat stream inside incubator, with temperature
distribution over incubator geometry (© MedIT, 2010)
mental conditions reach a state of balance. How-
ever, in reality, neither the baby’s heat production
nor the environmental conditions (Knobel, 2010;
Museux, 2004; Cloherty, 2008; Dodd, 2006).
3. MEDICAL INFRARED
THERMOGRAPHY
In the electromagnetic spectrum, infrared radia-
tion occupies the region between visible light and
microwaves. Generally, all objects in the universe
emit radiations in the IR region as a function of
their temperature. As an object gets hotter, it gives
off more intense IR radiation, and it radiates at
a shorter wavelength. The human eye cannot
“see” IR rays, but they can be detected by us-
ing IR cameras and detectors. The boundaries
between different IR spectral regions are not
agreed upon and can vary. In the following, the
boundaries that will be adopted in this work are
based on physical quantities of thermal radia-
tion. In general, IR radiation covers wavelengths
that range from 0.75 - 1000 µm, among which
the human body emissions that are traditionally
measured for diagnostic purposes only occupy
a narrow band at wavelengths of 8–12 µm. This
region is also referred as the long-wave infrared
radiation (LWIR) (Bissinger, 2010; Flenday, 2003;
Bohnhorst, 2010).
Historically, Sir William Herschel was the first
who recognized that heat emits in the infrared (IR)
wave spectrum (as early as in the 1800s). Medical
infrared, popularly known as IR-thermography,
and has utilized this heat signature since the 1960s
to measure and map skin temperatures pattern.
Our understanding of the regulation of skin
blood flow, heat transfers through the tissue layers,
and skin temperatures has radically changed during
these past 40 years, allowing us to better interpret
and evaluate these thermographic measurements.
During this same period of time, improved camera
sensitivity coupled with advances in focal plan
array (FPA) technology and new developments in
computerized embedded systems with assisted im-
age analysis have improved the quality of the non-
contact, noninvasive thermal map or thermogram
(Ginaliski, 2008). In the past decade numerous
93

application of infrared thermography were devel-
oped especially in medical and clinical diagnosis
field, as example of such application (early breast
cancer detection, rheumatoid diseases, peripheral
neuropathy, superficial vascular constriction re-
sponses, ophthalmic corneal diseases, and others)
(Ginaliski, 2007; Dodd, 2006; Vollmer, 2010).
This review will highlight some of the literature
and applications of thermography in medical and
physiological settings. More specifically, infrared
thermography and the structure and functions of
skin thermal microcirculation can provide a bet-
ter understanding of the following physiological
aspects:
(1) physiology of thermoregulation and skin
thermal patterns (e.g., comfort zone, influ-
ences of heat, cold, and exercise stressors,
thermoneutral zone, etc.).
(2) estimation of the skin blood perfusion rate
(e.g., traumatic and amputation site, skin
grafts).
(3) observation and diagnosis of the vascular
pathologies that manifest thermal distur-
bances in the cutaneous circulation.
(4) effect of thermal therapies evaluation.
(5) monitoring of the patient/subject/athlete’s
recovery as evidenced by the resumption
of normal thermal patterns during the reha-
bilitation process for some musculoskeletal
injuries.
Remember that an IR image is a visual map of
the skin surface temperature, which can provide
accurate thermal measurement, but it is poorly
quantify measurements of blood flow to the skin
tissue (Fortuna, 2010; Christidis, 2003). More-
over, it is also important to stress that recorded
skin temperatures may represent heat transferred
from within the body-core zone through various
tissue layers to the outer skin layer that may be
the result of conductive or radiant heat provided
from an external thermal stressor (Vollmer, 2010;
Fortuna, 2010 ; Gaussorgues, 1994).
94
Neonatal Infrared Thermography Monitoring
Accordingly, to interpret thermographic im-
ages, a basic understanding of physiological
mechanisms of bioheat transfer dynamics, skin
blood flow and factors that influence heat transfers
to the skin must be considered. With this under-
standing, objective data from IR thermography
can add valuable information and complement
other methodologies in the scientific inquiry and
medical practices.
3.1. Basic Principles of
Infrared Thermography
Generally, any object whose temperature is
above absolute zero Kelvin emits radiation at a
particular rate over a wavelength distribution.
This wavelength distribution is dependent on the
temperature of the object and its spectral emis-
sivity, ε(λ). The spectral emissivity, which may
also be considered as the radiation efficiency at a
given wavelength, is in turn characterized by the
radiation emission efficiency based on whether
the body is a black body, grey body, or selective
radiators. The black body is an ideal body, and
it is a perfect absorber that absorbs all incident
radiation and is conversely a perfect radiator. This
means that a black body absorbs and emits energy
to the maximum that is theoretically possible at
a given temperature. Therefore, within a given
wavelength (λ):
• ε = 1 for black body
• ε = constant < 1 for grey body
• 0 ≤ ε ≤ 1 for selective radiator
The radiative power W (or energy) and its
wavelength distribution is given by Planck´s ra-
diation law (Fortuna, 2010; Gaussorgues, 1994;
Nimah, 2006; Caniou, 1999; Teich, 1996):
2πhc 2   hc  
−1
W (λ,T ) =  exp  − 1 [Wm −2m −1 ]
λ5   λkT  
 
(7)
Neonatal Infrared Thermography Monitoring

Figure 8. The electromagnetic spectrum with illustration of infrared radiation classification and cor-
responding wavelengths (Modified from: Infrared vision, http://mivim.gel.ulaval.ca)

and the invisible infrared radiation emitted by an

object and convert it to a monochrome or multi-
colored image on a monitor screen wherein the
−1
2πc 2   hc  
P (λ,T ) = exp  − 1 [photons −1m −2m −1 ]
λ5   λkT   various shades or colors represent the thermal
 
(8) patterns across the surface of the object (Fortuna,
2010; Christidis, 2003).
where Essentially, the thermal imager may be coupled
h (Planck’s constant) = 6.6256 x10−34 J s with computer software for post-processing and
c (velocity of light in vacuum) = 2.9979 image analysis. Temperature readings may also
x108 msec−1 be registered. In general, thermal imagers of-
k (Boltzmann’s constant) = 1.38054 fer an excellent means of making a qualitative
x10−23 W sec K−1 determination of surface temperature, but there
λ = wavelength in μm remain many difficulties in obtaining an absolute
T = temperature in K. measurement.

The description of the electromagnetic spec- 3.2. Target’s Surface

trum and the classification of infrared wavelengths
spectral bands are illustrated in Figure 8.
While infrared radiation is invisible to the
human eye, it can be detected and displayed by
special IR thermal cameras. These cameras detect
Radiation Heat Exchange
The measurement of infrared (IR) radiation is the
basis for non-contact temperature measurement
and thermography. Thermal IR radiation leaving

95
 Neonatal Infrared Thermography Monitoring

Figure 9. Radiative heat flow mechanism (a), impinging process of radiation (b) on a target surface as the
element contributing to this process (1) total incoming energy,(2)radiation source, (3) absorbed energy,
(4) reflected energy,(5) transmitted energy,(6) total detected energy,(7) material properties (absorbance
(α), transmittance(τ) and reflectance (ρ)) and (8)infrared energy detector (camera) (© MedIT, 2010)

a surface (W) is called exitance or radiosity. It
can be emitted from the surface, reflected off the
surface, or transmitted through the surface. This
is illustrated in Figure 9(a). The total radiosity is
equal to the sum of the emitted energy fraction
(We), the reflected energy fraction (Wr) and the
transmitted energy fraction (Wt). The surface
temperature is related to We, the emitted com-
ponent only (Gaussorgues, 1994; Johnson, 1991;
Teich, 1996).
Generally, the thermal IR radiation impinging
on the surface can be absorbed, reflected or trans-
mitted as illustrated in Figure 9(b). Kirchhoff’s
law states that the sum of the three components
is always equal to the received radiation (the
percentage sum of the three components equals
unity)
Α(Absorbptivity)+ρ(Reflectivity)+
τ(Transmissivity)=1 (9)
Therefore, the sum of these components should
equal to unity, and each part of these physical
quantities has a value less than one (Johnson,
1991). For further simplification, if the infrared
thermal energy detector positioned in front of
the target surface, then the detected net radiation
(Wnet), is equal only to the transmitted component
of the infrared radiation.
Wnet = Wτ = τ ⋅WT (10)
This means, the other two physical quantities
(absorbed and reflected radiation) will not be
represented in the infrared thermal imaging and
this issue must be considered when dealing with
thermography calibration.
3.3. Specular and Diffuse Surfaces
Roughness and surface characteristics will influ-
ence the amount and direction of reflected radia-
tion. An ideally smooth surface will reflect incident
energy at an angle complementary to the angle
of incidence. This is called a specular reflector.
A totally rough or structured surface will scatter
or disperse some of the incident radiation. This
is called a diffuse reflector. In reality, all surfaces
have some diffusivity and some specularity. The
specular or diffusing characteristics of a surface
are taken into account by accounting for the emis-
sivity of the surface (Johnson, 1991).

96
Neonatal Infrared Thermography Monitoring
Figure 10. The black body radiation response over the temperature ranges within the specific wavelength
and according to Planck’s radiation law that describe the black body radiation (Dodd, Lancaster, Craig,
Smyth, & Williamson, 2006)
4. IMAGING PARAMETERS IN
INFRARED THERMOGRAPHY
Unlike most diagnostic modalities, medical ther-
mography is a non-invasive, non-contact imaging
method, and based on thermal radiation detection
within the body (Figure 10). Therefore, it is a
very sensitive and reliable means for graphically
mapping and displaying skin surface temperature.
The ability of a thermal camera to preserve
fine temperature details in the presence of large
scene temperature range is determined by its dy-
namic range (Nimah, 2006). Generally, this dy-
namic range is determined by the thermal camera’s
image digitization and formation electronics.
Further consideration should be taking when using
a thermal camera that allocates an adequate num-
ber of bits to the digitization of the images. Most
commercially available cameras use 12-bits or
more per pixel. This is quite adequate to preserve
fine details in images of the human body.
However, during image acquisition, the
camera field of view (FOV) should be empty
of any object that is dramatically cooler or
hotter than the subject; that is, avoid scene
temperature differences of more than roughly
30 °C (e.g., lamps, refrigerators, or radiators in
the background could cause trouble) (Nimah,
2006; Hyman, 2006).
This is analogous trying to use digital camera
to capture a picture of a person standing next to
headlights electronic circuits may bloom or sac-
rifice detail of the scene near the bright lights.
Although, the 12-bit digitization should preserve
fine temperature differences (∆T) at 30 °C, large
temperature differences generally stress the im-
age formation circuitry, and undesired artifacts
may appear. Nevertheless, it is relatively easy to
design a collection environment with a modest
temperature range. A simple way to do this is
to fill the camera field of view with the human
subject (Hanssler, 1992; Nimah, 2006).
97

4.1. Resolution and Sensitivity
The two most important parameters for thermal
camera sensors are sensitivity and resolution. The
thermal sensitivity is measured in degrees Celsius.
The modest thermal sensitivity is on the order of a
tenth of a degree Celsius. Good sensitivity sensors
can detect temperature differences up to 4 times
lower or 0.025 °C. Subsequently, this thermal
sensitivity is deemed valuable for medical diag-
nosis, since local temperature variations caused
by tumors and angiogenesis are usually higher
than this. The temperature resolution is analogous
to the number of colors in a computer display or
color photograph. The better the resolution, the
smoother the temperature transitions will exist
(Hyman, 2006; Caniou, 1999).
Essentially, if the subject has sudden tempera-
ture difference (∆T), those will be attributable
to the subject and not to the camera. The spatial
resolution of the thermal camera sensor-matrix or
what called focal plane array (FPA) is determined
primarily by the size of the imaging chip or pixel
count. This parameter is exactly analogous to the
world of proliferating digital photography. Similar
to the four megapixel digital camera can make
sharper photos than a two megapixel camera; pixel
count is a key element in the design of a medical
camera (Miles, 1991). There are quite economical
thermal cameras on the market with 320 × 240
pixels, and the images from such cameras can be
quite adequate for informal screening; imagery
may appear to be grainy if magnified unless the
viewing area or field of view (FOV) is reduced.
For example, if the image is of the full chest area,
about 18 inches, then a 320 pixel camera will
provide the ability to resolve spatial features of
about a sixteenth of an inch. If only the left breast is
imaged, spatial features as low as 1/32 inch can be
resolved (Miles, 1991; Caniou, 1999). On the other
hand, a 640×480 pixel resolution thermal camera
can cut these features sizes in half. Accordingly,
the good sensitivity and pixel count ensures that
the medical images will contain useful thermal
98
Neonatal Infrared Thermography Monitoring
and spatial detail. In summary, although 320×240
pixel resolution imagery is quite adequate, larger
pixel counts can provide more freedom for casual
use, and are essential for research and development
in medical centers (Caniou, 1999).
4.2. Thermal Imager Calibration
A further key feature for thermography imaging
is the temperature calibration. Fundamentally,
many thermal imaging systems are designed to
detect temperature differences, not to map cali-
brated temperature. A camera that maps the actual
surface temperature is a radiographic sensor. For
a precise temperature measurement, like imaging
a neonate inside a convective incubator, the IR
camera calibration will greatly deteriorate with
incubator’s temperature variation and humidity
condition. Therefore, the operator should map
local temperature differences over a small infant’s
anatomy with prior calibration. This application
would amount to a third eye for the physicians,
aiding them in finding sudden temperature gra-
dient and temperature anomalies within neonate
(Pollina, 2006; Heny, 2003).
It is critical that the calibration is stable and
accurate enough to match the temperature sensitiv-
ity of the camera. Here caution is advised. Many
radiometric cameras on the market are designed
for industrial applications where large temperature
differences are expected and the temperature of
the object is well over 100 °C; for example, the
temperature difference may be 5 °C at 600 °C.
In contrast to the calibration effect on medical
thermography (e.g., breast cancer), the tempera-
ture differences of interest are about a tenth of a
degree at about 37 °C. Therefore, the calibration
method must be suitable for such temperature
ranges (Teich, 1996). Since the dynamic range
of the breast cancer application is very small, the
calibration method is simplified. More important
are the temporal stability, temperature resolution,
and accuracy of the calibration. Useful calibra-
tion parameters are: of 0.1 °C resolution at 37 °C,
Neonatal Infrared Thermography Monitoring
Figure 11. Example of single band infrared thermal imager (MWIR) and (LWIR) systems (Courtesy of
InfraTec GmbH)
stability of 0.1 °C/h (drift), and accuracy of ± 0.3
°C. This means that the camera can measure a
temperature difference of 0.1 °C with an accuracy
of ± 0.3 °C at body temperature. For example,
suppose the breast is at 36.5 °C; the camera might
read 36.7 °C (Derflinger, 2009; Heidelise, 1998).
4.3. Single Band Infrared
Thermal Imagers
In general, there are two distinct spectral bands that
provide adequate thermal sensitivity for medical
use: First, is the mid-wave infrared band (MWIR)
which covers the electromagnetic spectrum from 3
to 5 μm in wavelength, approximately. Secondly,
is the long wave infrared band (LWIR), which
covers the wavelength spectrum from 7 μm to 13.5
μm. Typically, there are advocates for both bands,
and neither band offers a clear advantage over the
other for medical applications, although the LWIR
is rapidly becoming the most economical sensor
technology. Some experimenters believe that there
is merit to using both bands. Figure 11, illustrates
the two model of single band IR camera system, for
LWIR and MWIR respectively, were these types of
IR camera are widely used in different military and
civil applications (Saxena, 2008; Gerber, 2006).
The MWIR cameras are widely available and
generally have more pixels count, hence higher
resolution for the same price. In addition, the
MWIR sensors-array must be cooled to cryo-
genic temperatures as low as -196.15 °C in range;
therefore, it will increase the cost of such thermal
imagers (Saxena, 2008; Huang, 1999; Gerber,
2006). The phenomenology in this spectral band
has been quite effective in detecting small tumors
and temperature asymmetries within the human
body, accordingly, the MWIR-thermal imager
considered as the best choice for high-resolution
medical thermography application.
Thermoelectric coolers are used for some
MWIR sensors; they operate at temperature
(-98.15 °C to -53.15 °C) depending on the design
of the IR detector chip. In fact, the MWIR sensors
not only respond to emitted radiation from thermal
sources, but they also sense thermal radiation
energy from broadband visible sources such as
the sun. Particularly, the images in this spectral
band can contain structure caused by reflected
light rather than actual emitted radiation. Table 1
gives a summary of the important parameters for
the infrared thermal camera systems.
Consideration must be taken to minimize re-
flected light from broadband sources including
99
 Neonatal Infrared Thermography Monitoring

Table 1. Summary of the main and important parameters for the infrared thermal camera systems

Summary of important infrared thermal camera parameters

Parameters Application: recording Application: informal
Format Digital stills Video or stills-image
Compression None As provided by mfr (mean frame
rate)
Digitization (dynamic range) 12 bits or more 12 bits nominal
Pixels (array size) 320 x 240 up to 640 x 480 320 x 240
Sensitivity @ 30°C 0.04 °C, 0.1°C max. 0.08-0.1 °C
Calibration accuracy ± 0.3 °C Not required
Calibration range Room and body temperature Not required
Calibration resolution 0.1 °C Not required
Spectral band MWIR or LWIR MWIR or LWIR
Data interface USB -IEEE 1394 Firewire S-video, IEEE 1394 Firewire

incandescent light bulbs and sunlight. Unwanted
light can cause shadows, reflections, and bright
spots in the imagery. Basically, direct illumination
of the subject by wideband artificial sources and
sunlight should be avoided. Moreover, lighting
geometries and sources should be shut down
before thermography measurement. As well as,
moisturizing creams, sweat, and other skin surface
coatings should also be avoided (Saxena, 2008;
Huang, 1999; Browne, 2004).
5. NEONATAL TEMPERATURE
MONITORING TECHNIQUE
5.1. Skin Temperature Measurement
In this section we will lighten the concept of neo-
natal monitoring and the enlightenment of need
for non-contact temperature observation inside
NICU stations. Principally, the skin temperature
is a vital physiological measurement, and its ac-
curacy is limited by variability in the character-
istics of neonate, transducers-sensor, and neonate
care techniques. Basically, the neonate’s surface
temperatures are not homogeneous because of
difference in following aspects:
(1) skin and subcutaneous tissue thickness over
different body parts
(2) differences in structures underlying different
skin surfaces
(3) variation in the vascularity (tissue vascu-
larization) and neurovascular response over
body regions.
Therefore, different skin surfaces have dif-
ferent temperatures, and when temperature
sensors are connected to the skin surface, they
measure the temperature only at the specific site
(localized temperature mapping). Moreover,
the thermistors are attached using devices that
caused compression of superficial skin vessels
underlying the thermistor element. Thus, by their
very attachment, thermistors modify both the
absolute temperature they are expected to mea-
sure and the spontaneous dynamic variability in
the measured skin temperature that is normally
affected by changing amounts of warm blood
flowing through the skin over different time
periods. Since temperature should be monitored
in all neonates undergoing different clinical and
surgical procedures, therefore, thermistor will
not provide precise temperature measurement
within daily care.

100
Neonatal Infrared Thermography Monitoring
Figure 12. Four different types of temperature electrodes used in the daily clinical care and monitoring
inside NICU wards (a) rectal type thermistor, (b) skin type thermistor, (c) axilla type thermistor, (d) foot
thermistor (© MedIT, 2010)
Basically, the core temperature can be esti-
mated at several sites of transluminal cavities,
including esophagus, rectum, tympanic membrane
or nasopharynx, conversely, core temperature can
be estimated over forehead or over supranasal
region (Fransson, 2005). Esophageal probes are
the most convenient and can be incorporated into
an esophageal stethoscope. In addition, the skin
temperature should also be measured because the
gradient between core and peripheral temperature
gives a useful indication of the adequacy of cardiac
output. Several types of these medical thermistors
illustrated in Figure 12
5.2. Non-invasive Skin Temperature
Monitoring
During the inception of neonatology as a specialty,
the physiological monitoring in the neonatal inten-
sive care unit (NICU) has improved dramatically.
In fact, several devices and modules are increas-
ingly relied on to alert the physician to developing
problems and even to direct care as technological
advances, and make the newborn infants man-
agement inside NICU more complex. Although,
the clinical monitoring systems are diverse, they
should all meet a few basic requirements, which
set according to the international electrical com-
mission for medical equipment standards of infant
incubator (IEC 60601-2-19 2009) and their related
addendum (Fransson, 2005). These requirements
are as follows:
• Reliability, the system should be reliable,
both in the sensitivity and the specificity of
the information relayed by the system and
in long-term equipment integrity.
• Simplicity, it should be relatively simple
to operate and provide information in an
easy-to-interpret manner.
• System safety and hygienic aspect, the sys-
tem must be safe for patient use, which is
especially important in the neonatal popu-
lation because the sensors or probes, the
101

portion of the system that comes in con-
tact with the patient, must be appropriately
sized and nonirritating to sensitive skin.
• Non-invasive technique, the ideal system
should be non-invasive, or at least not re-
quire invasive procedures beyond those
which are routinely needed for optimal pa-
tient care.
• Real-time operation, the system should
provide continuous, or near-continuous,
real-time information, such that personnel
may respond to an event occurring at the
time the information is relayed. In addi-
tion, it should be relatively small and por-
table (Fransson, 2005; Fanaroff, 2007).
As general, the system must be safe, and must
not interfere with other vital equipment required
for patient care. Thus, according to criteria, the
neonatal monitoring system should be as non-
invasive as possible to avoid any stressing to the
neonate. From this point the application of the
infrared thermal imaging to mentoring aspect in the
NICU. The use of infrared thermal camera within
the incubator has a high safety factor when com-
pared to other invasive measurement technique,
such as rectal and axillaries temperature sensors,
pulse oximetry sensor...etc) (Bhatia, 1976).
Therefore, developing an accurate and efficient
monitoring technology for daily neonatal care will
increase the survival rate of newborn infants. For
instance, one of the best solution for the contact-
less vital signal monitoring is the infrared thermal
imaging, by which a registration and detection of
different heat-generated physiology (e.g., super-
ficial vascular blood flow during thermal stress,
air-conditioning of the inspired air, temperature
variation over the skin, blood perfusion), thus,
within the neonatal clinical observation, stand the
temperature as the important one. With remarks
to all above clinical requirement of temperature
monitoring, a contactless measurement method
based on infrared thermography imaging, was
applied inside NICU ward. Moreover, this tech-
102
Neonatal Infrared Thermography Monitoring
nique and its related application will be discussed
deeply in the next sections.
6. EXPERIMENTAL SETUP FOR
NEONATAL INFRARED IMAGING
Recently, at Aachen University Hospital, the
method presented in this chapter was implemented
by using infrared thermography images acquired
from VarioCam hr-Head (InfraTec GmbH, Ger-
many) thermal camera, which is long wave IR-
FPA bolometric detector, with (7-14µm) spectral
bandwidth. The camera transferred the thermal
observations to a remote laptop via IEEE-1394
Firewire interface. The original research reported
in this chapter has been approved by the local
medical ethics committee on 19.08.2009, running
number (EK032/09).
The neonate’s thermal images inside incubator
or other open-neonatal care system (e.g., pho-
totherapy system and infrared radiant warmer)
were converted to thermal video data. Addition-
ally, these data were used to test the algorithmic
software for tracking specified regions of interest
(rois) on the neonate’s skin (forming a virtual
temperature sensor) (Fanaroff, 2007). In general,
the schematics of the thermal ir-camera setting as
a typical experimental setting for neonatal infrared
thermography (nirt) displayed in figure 13.
The proposed configuration totally differs from
the first infrared measurement setting, used by
Adams et al. (Clark, 1980; Chadron, 2006), in
which he made a hole in the incubator surface
with diameter of 45 cm in order to visualize the
baby inside incubator with aid of infrared-mirror
(as reflecting surface). Thus, he modified the
physical properties of the incubator (i.e., convec-
tive heat loss increases from incubator). At the
moment, this technique would not be still appli-
cable within the medical ethics regulation of
neonatal intensive care unit; hence, avoiding such
solution is the main aim of this clinical measure-
ment (Bhatia, 1976; Kent, 2008).
Neonatal Infrared Thermography Monitoring
Figure 13. The current setup for neonatal infrared thermography used by the authors inside the NICU
with all attached cables and monitoring terminals (© MedIT, 2010)
Primarily, the contact temperature electrodes
attached to the neonate body in two sites. The
first, is the back region were this region is almost
exposed to the convective air stream inside incu-
bator, and other electrode attached to the belly
region (e.g., over the liver) in which it reflects
the body core temperature of the neonate’s body.
These two electrode play as reference temperature
measurement for the NIRT imaging procedure, to
compared with infrared thermal data. Therefore,
any changes in the incubator physical properties
are avoided (Kent, 2008; Horns, 2003).
Essentially, IR-thermal camera calibration
is necessary during clinical measurement inside
NICU, in order to avoid any thermograms qual-
ity deterioration through measurement process.
Basically, this deterioration caused by tempera-
ture and humidity variation inside incubator, and
the calibration process is called Automatic Non
Uniformity Correction (ANUC). This calibration
process implemented during the clinical measure-
ment using the IRBIS® Professional software of
the IR camera. Initially, the selections of object
of interest (OOI) inside thermogram; have to be
specified first, and the setting the environmental,
incubator and the object, through IR transparent
window made with (0.01 mm polyethylene (PE)
material). Principally, the transmissivity of IR
(PE-foil) is between 0.91-0.94. This transparent
foil closed the clapper opening to visualize the
baby inside the incubator, because the Plexiglas
material of the incubator hood is IR-reflected
material with emissivity value reach (0.97).
Therefore, it hides the neonate from the thermal
camera.
6.1. Effect of Viewing Angle
on Neonatal Thermography
Viewing angle is an important point in infrared
thermography imaging, because it defines the
actual field of view required to cover large surface
area of the object. Essentially, the target size can be
estimated from the viewing angle as shown below:
103
 Neonatal Infrared Thermography Monitoring

Figure 14. Schematic of the viewing angle variation for neonatal infrared thermography imaging where,
this viewing angle should cover the whole neonate’s body as possible as the observer can (© MEDIT, 2010)

Target size = θresolution ⋅ R, (11) Therefore, the vertical field-of-view (VFOVFootprint)

footprint is estimated as in the following equation:
where θresolution is the IR camera depression angle,
 
and R is the distance to the target (newborn infant), 



1 1
assuming that the target is normal (perpendicular) to VFOVFootprint =h − ,
  VFOV   VFOV  
 tan θ −  tan  θ + 
the line-of-sight. However, this is not always the case.   2 

 2  
When observing a roof from a distance, you are view- (14)
ing it at an oblique angle; this will distort the pro-
jected foot print. Figure 14 illustrates the viewing The viewing angle condition is essential for cor-
geometry when looking downward to a neonate inside rect NIRT measurement, because it is important
incubator. This typifies a roof surface where the ob- to know the apparent VFOV for the tested target
server is standing and using a handheld IR camera. (newborn infant). This is because the viewing angle
In the center of the horizontal field-of-view, will decide how much the coverage of target’s area
the distance to the target at the bottom and the top (Figure 15). Therefore, viewing angle modifica-
of the field of view (FOV) (when looking down- tion should perform before the measurement start.
wards) (Rbuttom and Rup respectively) is computed
as follows:
7. NEONATAL’S TEMPERATURE
R bottom =
h
(12) DISTRIBUTION MAPPING
 VFOV 
sin θ + 
 2  The heat exchange between the infant’s skin and
the environment is influenced by the insulation
h provided by the skin, the permeability of the skin,
R up = (13) and environmental factors such as the ambient
 VFOV 
sin θ −  temperature and humidity, air-flow velocity, and
 2 
the temperature and characteristics of the incu-

104
Neonatal Infrared Thermography Monitoring
Figure 15. Variation in viewing angle for NIRT imaging will lead to varying surface area coverage of
target surface (newborn infant) as thermography frames (a), (b) and (c)
bator surfaces facing the infant. Basically, the
evaporative heat loss from the skin is the major
component of heat exchange in the most preterm
infants early after birth, and these infants gain heat
through convection and, in a very dry environ-
ment, possibly also through radiation when they
are nursed inside incubator.
As water loss from the skin surface of the
most preterm infants’ decreases with postnatal
age, heat loss through evaporation from the skin
also decreases. Their need for a high ambient
temperature also diminishes, and with the lower
temperature of the incubator walls, the heat loss
through radiation then increases and the heat gain
by convection is changed to a low loss of heat
during the first weeks after birth (Merenstein,
1998; Jamsa, 1998).
To demonstrate the capabilities of IR thermog-
raphy, an experiment performed to quantify the
effect of kangarooing, i.e., to measure temperature
distribution on the neonate while taken out of the
incubator and placed on the mother´s upper torso.
Basically, the proposed protocol for neonatal IR
thermography (NIRT) measurements composed
of four phases, and each of these phases with 10
minutes time duration (Figure 16).
Throughout NIRT measurement, one can
observe the variability of the surface temperature
distribution over different anatomical segments.
Basically, this is due to different physiological
reactions of these segments, as we can see in
Figure 17, in which we have two different thermal
images taken at the start of the measurement and
at the end of the measurement.
Obviously, we can see the temperature migra-
tion patterns from the different segments like
(hands, arms) to the central area of the body (core).
This effect as hypothetically explained may due
to several factors one of them is the metabolic
reaction rate of the neonate responds to the ex-
ternal thermal stresses or interpretation by the
nursing staff or the mother interaction (Merenstein,
1998).
8. PREDICTION OF HYPOTHERMIA
OR HYPERTHERMIC CONDITIONS
Primarily, most of neonatal thermoregulatory
studies focused on the effects, prevention, and
amelioration of hypothermia. Hyperthermia has
been noted mainly as a sign of hypermetabolism
when an infant is septic or otherwise stimulated.
It is probably beneficial to cool an infant who has
become febrile because of exposure to an over-
heated environment. Whether it is good to cool
infants who are febrile because they are septic, or
otherwise stressed by internal conditions, is less
clear, although it is usually attempted (Jamsa,
1998; Arora, 2008). In general, an infant with a
body temperature higher than 37.5 °C is often
considered to be abnormally warm. To determine
105
 Neonatal Infrared Thermography Monitoring

Figure 16. Neonatal infrared thermography (NIRT) protocol used in the measurement of the surface skin
temperature for the neonate under different clinical conditions (© MedIT, 2010)

whether the elevated temperature is caused by an
increase in heat production, which might occur if
the infant is septic, or by a decrease in heat loss,
some simple presumptive clinical measures can
be made.
A physiologically competent infant responds
to overheating from a hot environment by incor-
porating heat-loss mechanisms. The infant’s skin
vessels dilate, the infant may appear flushed, the
hands and feet are suffused and warm, and the
infant assumes a spread-eagle posture. Evaporative
losses increase and, although it is uncommonly
observed in premature infants, active sweating may
be noted in the full-term infant. If the heat stress
is severe, the infant may become hyperactive and
irritable. During rapid warming, the skin of the

Figure 17. Two thermal images of a neonate inside a convective incubator; which shows a very slight
variation of surface temperature overall neonate’s anatomical regions (© MedIT, 2010)

106
Neonatal Infrared Thermography Monitoring

infant is warmer than the infant’s core temperature
(Arora, 2008).
During the NIRT measurement, the estima-
tion of the hypothermia (decreasing of the body
temperature) and hyperthermia (increasing the
body temperature) can be performed through
a serial registration of the defined ROIs within
specific anatomical segment (e.g., back, belly and
extremities) for the definition of the temperature-
time plotting and evaluating the degree of the
temperature gradient, whenever, this gradient indi-
cating serious-to-moderate temperature increase or
decrease. Figure 19 indicating the plotting of these
skin surface temperature profile (Table 2) over 10
minute recording inside convective incubator for
two different newborn infant. Note that from the
measurement illustrated in Figure 19, it is quite
obvious that the neonate thermoregulation is not
a passive process.
The temperature measurement of neonate
performed according to specified point over in-
fant’s body (Figure 18). Principally, these points
specified within zones according to the anatomi-
cal distribution of temperature in reference to the
previous studies. Generally, the number of these
zones are five, but actually it can be increased
when more area of interest need to evaluate using
infrared thermal imaging. Moreover, from these
location and corresponding temperature measure-
ment, the neonatologists and infant care personnel
can predict the signs of early hypo/hyperthermia
and related metabolic disturbances. Therefore,
the temperature patterns changes can be detected
very early, when considering these anatomical
zones as referential territories for NIRT imaging.
Hence, the temperature changes throughout NIRT
measurement is varied over these zones and this
can be clearly seen in Figure 19, where these
points (P1…P5) registered their temperature
trends for 10 minute time interval as previously
explained in NIRT measurement protocol.
9. APPLICATION EXAMPLES
In the following, three clinical examples will be
provided to illustrate the potential of possible
NIRT imaging applications. These clinical ap-
plications, varies from temperature distribution
mapping of neonate under kangaroo mother care
(as open-therapy type), phototherapy monitoring,
breathing rate monitoring by detecting infrared

Table 2. Illustration of the infrared thermography measurement over specified points compared with
skin electrode measurement for incubator phase 1with corresponding point assignment as illustrated
in Figure 15

Comparision of infrared temperature over specified ROIs of neonate’s body vs. skin temperature electrode (incubator phase 1)
baby P1 P2 P3 P4 P5 ECDback ECDbelly
1 34.03 35.12 35.34 36.23 36.19 36.5 37.02
2 34.35 35.20 36.03 36.16 36.20 36.7 37.13
3 34.17 35.67 36.11 36.27 36.08 37.1 36.80
4 34.27 35.41 35.88 36.56 36.28 36.5 36.75
5 34.21 35.84 35.87 36.42 36.90 36.6 37.10
6 34.08 35.40 36.07 36.40 37.04 37.0 37.12
7 34.33 35.82 36.11 36.42 36.82 36.6 37.02
8 34.78 35.67 36.09 36.31 36.27 36.2 37.00
9 34.12 35.93 35.85 36.43 36.49 36.7 37.30
10 34.36 35.46 35.91 36.59 36.61 36.6 36.95

107
 Neonatal Infrared Thermography Monitoring

Figure 18. Example of the point locations for NIRT measurement, which considered as standards posi-
tion for incubator phases one and two (© MedIT, 2010)

Figure 19. Two thermograms of neonates inside a convective incubator for two different babies; the
right temperature trends of 4 points shows variation of the surface temperature over different anatomi-
cal segments (© MedIT, 2010)

108
Neonatal Infrared Thermography Monitoring
thermal signature of respiration activity over
new-born infant nostril’s region. Essentially,
these promoted applications will be the basis for
the future development of smart and wireless
incubator technology.
9.1. Example 1: Monitoring
Kangaroo Mother Care with
Infrared Thermography
Kangaroo mother care (KMC) is defined as the
care of preterm infants carried skin-to-skin con-
tact with the mother. It is consider as a powerful,
easy-to-use method to promote the health and
well-being of infants born preterm as well as
full-term. Throughout different studies on the
thermoregulation effect of KMC, shows that,
the KMC will bring thermoregulation function
of neonate to better adaptability. Actually, this
adaptability may refer to the direct interaction
between mothers and neonate’s skin temperatures.
Although, the conduction heat transfer will be ma-
jor component for KMC, but there will be radiative
heat transfer between mother and neonate (Arora,
2008; Gibbson, 1997). With the basic feature of
the KMC can be summarized as follows:
• Skin-to-skin Contact, which is between
the baby front and the mother’s chest. The
more skin-to-skin, the better for comfort a
small nappy is fine, and for warmth a cap
may be used. Skin-to-skin contact should
ideally start at birth, but is helpful at any
time.
• Psychological dyad between the mother
and the baby makes the inter-thermoregu-
lation process more optimal and providing
surface temperature stability over kanga-
rooing time.
After introduction the skin-to-skin care for
newborn infants, Fanaroff et al. (Fanaroff, 2007;
Ibe, 2004) showed that preterm infants less than
one week of age and with a birth weight of less
than 1500 g were not exposed to cold stress dur-
ing 60 minutes of such care. Similarly, preterm
infants born at 28 to 30 weeks of gestation and
studied during the first and second week after
birth increased their body temperature during 1
hour of skin-to-skin contact with no significant
change in oxygen consumption (Gibbson, 1997;
Kent & Williams, 2008).
Basically, the kangaroo mother care (KMC)
is a part of the overall clinical measurement
for neonatal thermography, where the KMC
procedure including in the central part of the
NIRT clinical measurements. Principally, the
kangaroo mother care (KMC) measurement
setup, consist of two targets, first is the mother
(or in some case father) holding the neonate
(second target) on her/his chest (Figure 20).
Initially, the covering of the neonate surface
perform by IR transparent thermal isolated
polyethylene (PE) foil. This foil will provide
visibility to infrared thermal camera used to
acquire real-time thermal image during the
KMC procedure (Figure 21). Moreover, the
measurement setup including the registration of
different physiological parameters (e.g., heart
rate, oxygen saturation, respiration rate, etc)
(Bohnhorst, 2001; Conlon, 2004).
Generally, the kangaroo mother care (KMC)
has a collateral thermal effect produced by the
convective air-stream of the ventilated air inside
intensive station room. This effect should also
consider during the final estimation of thermo-
regulatory patterns of neonates under KMC
procedure (Bosque, 1995; Ibe, 2004).
For more emphasizing on skin regional mea-
surement, proper regions of interest (ROI) located
for the kangaroo mother care (KMC), in which
these regions are used to measure the temperature
patterns over neonate body. Basically, these points
are varied from infant to infant, although, all of
them should be located inside the defined areas
(Figure 21).
Seemingly, the data collected from KMC mea-
surement tends to have as low oscillation pattern
109
 Neonatal Infrared Thermography Monitoring

Figure 20. Panoramic view of the kangaroo mother care procedure inside the neonatal intensive care
unit (NICU), this setup is also approved as clinical setup configuration (© MedIT, 2010)

Figure 21. Defined ROIs for infrared thermography used in kangaroo mother care (KMC), (© MedIT, 2010)

110
Neonatal Infrared Thermography Monitoring
than the incubator phases temperature pattern, this
may be due to the less heat-energy loss from the ne-
onate’s body and the forced thermoregulation from
the mother side. Accordingly, the illustration of the
temperature profile of the KMC over a specified
point defined, showing stable temperature pattern,
despite the descending or ascending temperature
value throughout the KMC measurement (Figure
22). The measurement temperature scenarios of
the KMC phase also show variable pattern for
different neonate’s anatomical segments, as a
process of thermoregulation dynamics of the baby
and mother. Principally, the underlying process of
thermoregulation of neonate is unknown and what
exactly happening between mother skin surface
and neonate’s skin is unknown also.
Actually, one hypothesis believed to be true for
the KMC care procedure, which is the conductive
heat transfer route, which becomes the primary
heat balance for neonate. This heat balance comes
with counter current heat flow of mother-baby
pathway. As basic knowledge, that the adult human
is more robust to the wide range of temperature
change (Gibbons, 1997; Dzukou, 2004; Conlon &
Drolet, 2004; Merenstein & Gardener, 1998).
Figure 22. The contouring plotting of two infrared thermography imaging for kangaroo mother care
(KMC) at (a) start and (b) end of measurement (© MedIT, 2010)
9.2. Example 2: Monitoring the
Effects of Neonatal Phototherapy
Basically, the neonatal jaundice result from the
accumulation of bilirubin in the body at concen-
trations greater than can be handled by the liver.
It is a common medical condition that affects
65% of newborns. The bilirubin occurs naturally
in the body as red blood cells are broken down.
Bilirubin begins as fat soluble, or un-conjugated,
until it is transported to the liver by albumin
and converted to water-soluble, or conjugated,
bilirubin. The water-soluble bilirubin is easily
excreted through urine, stool, or sweat glands. At
birth the concentration of red blood cells is high,
and consequently the concentration of bilirubin
is high. Since bilirubin is largely broken down
in the liver, newborns with premature livers are
unable to handle high levels of bilirubin and of-
ten suffer from jaundice. Physical symptoms of
jaundice include yellowing of the eyes and the
inside of the mouth; conjugated bilirubin gathers
on the surface of the body causing these physi-
cal changes. The three common treatments are
employed for this condition: blood transfusion,
phototherapy, and medications (Hansen, 2000;
Thomas, 2001).
111

Figure 23. The imaging setup of neonatal infrared imaging under phototherapy procedure (© MedIT, 2010)
The phototherapy (PT) is defined as using the
visible light for the treatment of hyperbilirubi-
nemia pathology in the newborn. This relatively
common therapy lowers the serum bilirubin level
by transforming bilirubin into water-soluble iso-
mers that can be eliminated without conjugation
in the liver. The dose of phototherapy largely
determines how quickly it works; the dose, in
turn, is determined by the wavelength of the light,
the intensity of the light (irradiance), the distance
between the light and the infant, and the body
surface area exposed to the light. Commercially
available phototherapy systems include those
that deliver light via fluorescent bulbs, halogen
quartz lamps (HQL), light-emitting diodes (LED),
and fiberoptic mattresses. Proper nursing care
enhances the effectiveness of phototherapy and
minimizes complications associated with this
procedure (Knobel, 2011). Therefore, the care-
giver responsibilities include ensuring effective
irradiance delivery, maximizing skin exposure,
providing eye protection and eye care, care-
fully monitoring thermoregulation, maintaining
adequate hydration, promoting elimination, and
supporting parent-infant interaction (Eggert, 1984;
Hansen, 2000).
112
Neonatal Infrared Thermography Monitoring
Essentially, the phototherapy experimental
setup with thermography imaging is the same as of
NIRT setup but with incubator under photothera-
peutic source (Figure 23). The calibration process
for thermography is slightly different due to the
presence of external thermal source (Phototherapy
light bulb), which emits a sufficient amount of
energy, therefore, the thermography imaging must
compensate for this external heat source.
As example of phototherapy care infant under
infrared thermogram, noting that infrared ther-
mogram for the neonate under phototherapy
procedure is similar to the regular neonatal ther-
mography (Figure 24). Initially, the infrared
thermograms show no significant changes in
skin-surface temperature, makes such assumption
that the prolonged phototherapy does not make
any consequential hyperthermia effect on neonate.
The clinical measurements are performed in
cooperation with the neonatal intensive care station
at RWTH Aachen University Hospital. Addition-
ally, elaborate the first measurement protocol for
phototherapy study performed under infrared
thermography as sub-study of NIRT imaging.
Consequently, this measurement considered the
closed convective incubator therapy instead of
Neonatal Infrared Thermography Monitoring
Figure 24. Thermal images of the neonate under phototherapy (PT), (left) at the measurement starting
and (right) at the end of measurement, as appear through a temperature scale, there is no significant
change of temperature of neonate when undergoing phototherapy (© MedIT, 2010)
open-based therapy, in which there is a physical
barrier (incubator Plexiglas hood) that may absorb
the thermal energy developed by the light-bulb of
the PT device (Moseley, 1988; Owa, 2011).
Basically, this physical barrier which is in nature,
a good infrared radiation reflector, attenuate a great
portion of the thermal radiation developed from
the light bulb, which have radiation energy within
range of (300 Watt/m2 to 450 Watt/m2). Knowing
that the maximum radiation energy that reaches
the neonatal skin surface is between (20 µWatt/cm2
to 30 µWatt/cm2) (Bhutani, 2012; Eggert, 1984).
In summary, the phototherapy procedure has
no diverse effect on the neonatal skin temperature,
despite the high radiation energy emitted from
the phototherapy lamps. This finding will be the
foundation for further research to investigate
the effect of the PT procedure on the neonate’s
thermoregulation functions.
The modeling and simulation of the PT process,
with advent of thermography imaging, will let
to know more about the heat transfer dynamics
through the neonatal skin component.
9.3. Example 3: Non-Contact
Monitoring of Neonatal Breathing
Respiratory monitoring and assessment are es-
sential in the neonatal intensive care unit (NICU).
Although, most of the physical examination can
supply information regarding the quality of an
infant’s pulmonary effort, it is considered as a
noncontiguous evaluation (Kumar, 1996). There-
fore, the continuous monitoring is critical in the
sick patient, but the quantity of information var-
ies greatly with the used method. Basically, the
surface (contact) monitoring techniques are used
routinely in the NICU to monitor respiratory rate,
but they are unable to provide detailed information
regarding further pulmonary parameters and are
often subject to artifact (Pavlidis, 2007). Methods
of sensing ventilation requiring tight connections
with the patient’s airway often provide the most
data, but they are invasive and, until recently, were
impractical for continuous use. It is important for
all care providers to understand the benefits and
limitations of the respiratory monitoring systems
used in the care of their patients (Pavlidis, 2002,
2006, 2007; Sun, 2006).
Increasing demands for the non-contact moni-
toring solution inside the neonatal intensive care
unit, force the research to develop technology to
replace the conventional respiration monitoring
methods. The first approach for measuring the
breathing rate non-invasively with thermography
was done through the work of Palvidis et al. (2002,
2006, 2007) and Sun (2006) while he used the
exhaled carbon dioxide (CO2) gas from the nostrils
113
 Neonatal Infrared Thermography Monitoring

Figure 25. The image series of IRTR signal detection process in the neonate (a). The sequence indicating
the tracking of the one respiration cycle for 1100 msec (b), this interval is not fixed over the measure-
ment interval, but varies according to the breathing physiology (© MedIT, 2010)

against infrared reflected surface (e.g., Copper
(Cu) or Aluminum (Al) plate), and register these
CO2 jet during respiration cycles (inspiration and
expiration). Moreover, he also detect the blood
flow at superficial arteries and veins to quantify the
regional blood flow at the specific sites (Pavlidis,
2002, 2006; Sun, 2006; Zhu, 2007).
The vital signal monitoring based on IR ther-
mography is a new growing field of biomedical
engineering (Pavlidis, 2002, 2007; Garbey, 2007,
Sun, 2006). One important physiological signal of
interest in a clinical diagnosis is the respiration rate.
Basically, the respiration rate can be observed at
the anterior naris (nostrils), when the temperature
profile associated with inspiration and expiration
phases changed.
Recently, under the intensive research work
done with the collaboration of Philips chair for
medical information technology (MedIT) and
RWTH Aachen University Hospital, founded
that it is also possible to detect the temperature
variation due to the convective heat loss in the
nostrils region of the neonate, which is certainly
lower than that of the adult with temperature
difference (∆T) approaching 0.2 °C to 0.5 °C as
compared to the human adults with temperature
difference (∆T) approaching (2 °C to 4 °C) in
average (Abbas, 2011).
One of the example of such detected respiratory
thermal variation measured from nostrils region
of neonate (Figure 25), where the respiration rate
in this measurement approaching about 38 cycle/
minute, which reflects good detection criteria for
respiration rate (Abbas, 2009).
Basically, the setup of IRTR measurement
composed of thermal camera located at a distance

114
Neonatal Infrared Thermography Monitoring
Figure 26. The experimental setup of the infrared thermography respiration monitoring (IRTR), as a
measurement used to detect the heat convective loss through the respiration cycle, and register these
thermal deviations as temporal trends (© MedIT, 2010)
between (25 cm to 40 cm) in order to detect the
maximum nostrils temperature variation (Figure
26), and to assure a good calibration setting through
the measurement time. This setting is proposed
after examining the signal intensities over five
neonates to be the physically fit for this type of
thermography measurement. The main problems
which faces this measurement is the small thermal
signal intensity detected within the neonate res-
piration cycles, and the neonate’s face movement
which is normally occur in many neonatal phys-
iological measurement. The last problem is the
main cause of IRTR signal deterioration and we
develop a compensation algorithm for this effect
based on thermography tracking technique. This
method based on following the neonate’s face
gesture, and define nostril’s region to register the
temperature variation. The preliminary results
from IRTR experiments were very encouraged to
proceed with optimization of the measurement
protocols to be considered as a new breathing
monitoring technique inside NICU wards.
The IRTR signal measurement, include the
recording of the heart rate and the respiration rate
to use these measurements, as reference-base for
comparison and interpretation purposes. The
profile definition over the nostrils is not constant
for all neonate setting, this because the neonate
resting in different horizontal position, and sub-
jected to change this position over time. This
problem should encounter during the IRTR mea-
surement, because we cannot guarantee the detec-
tion of the nostrils thermal signature unless we
have overlay the camera field of view (FOV) over
the nostrils with minimum degree of (30°) (Figure
27). Alternatively, we can also monitor the respi-
ration thermal signature from the mouth when the
neonate is with continuous positive airway pres-
sure (CPAP) ventilation or respiration intubation
mask (Abbas, 2009, 2011).
115

Figure 27. The approximation of the IR camera setting for neonatal respiration detection and the region
of nostrils need to focused on for effective field of view (FOV) (© MedIT, 2010)
Despite that, this method has a good coverage
surface area-to-signal contrast (SA/SC), but it
also not register the direct relation of the respira-
tion thermal variation (Pavlidis, 2002, 2007).
Furthermore, to assured the IRTR signal detection,
a full coverage IR-reflected surfaces required to
enclose the face geometry through thermography
acquitsion. From medical safety point of view,
this solution is not practical one to implement
inside neonatal incubator. This because, the high
humidity inside incubator cause metallic surface
oxidation, therefore, the IR reflected surface
should be oxidation-free and non-toxic to avoid
any hazards to the new born babies.
Normally, the result acquired from the IRTR
measurement, shows a slight variation of the
nostrils temperature (Figure 28). Apparently, this
variation may originate from infrared thermal
camera calibration, as instant compensation for
temperature gradient inside convective incubator.
The respiration rate from both methods conven-
tional respiration detection form the ECG signal
and the IRTR method are listed as in Table 3.
116
Neonatal Infrared Thermography Monitoring
Although, this method have a promising ap-
plication to monitor respiration rate, but it lacks
also the robustness of breathing activity registration,
due to the motion artifact and the difference in
anatomical shapes of the babies nose, which need
the infrared camera to adapt for these changes.
Technically, the above mentioned problems can
be compensated by using computer-vision based
tracking algorithm, in which this algorithm fol-
lows the nostrils region in the successive frames
of thermography images, and then register the
heat gradient as pseudocolor contract changes
over time (Abbas, 2008, 2009)
10. OUTLOOK: POSSIBLE
DIRECTIONS OF FUTURE
RESEARCH
From the examples given so far, the authors believe
to have demonstrated that thermal IR imaging
has the potential to become a standard method of
monitoring in neonatal intensive care. Indeed, it
Neonatal Infrared Thermography Monitoring

Figure 28. Neonatal Infrared Thermography imaging (a) inside neonatal incubator with defined ROI
around the mouth opening, (b) fluctuating temperature-time profile of the neonate during normal daily
care activity (© MedIT, 2010)

may also have strong medical implications, such
as monitoring the effects of drug-medication
administration. Technically, it could make sense
to incorporate IR technology into incubators,
provided that the technology continues to become
cheaper.
A possible future design and setup of a next-
generation neonatal incubator, can integrate dif-
ferent medical imaging and therapeutic modalities
which are important to the neonate healthcare.
Basically, the main components for this incuba-
tor design consist of the following as illustrated

Table 3. Comparison of the respiratory rate derived from IRTR and from conventional ECG methods

Subj. RR (IRTR) (bpm) RR from ECG (bpm) Maximum ROI temp [°C] Minimum ROI temp [°C]
1 42.50 40.36 32.23 32.85
2 44.25 42.60 32.11 32.77
3 39.40 38.60 33.26 33.49
4 45.14 45.20 32.18 32.45
5 53.32 52.09 32.85 32.31

117
 Neonatal Infrared Thermography Monitoring

in Figure 29, (1) incubator hood (dome shaped)
made from Plexiglas material, (2) workstation
display, (3) ergonomic transportation base, (4)
infrared thermal camera, (5) SpO2 camera, (6)
neonate cared inside incubator, (7) integrative
humidifier system (new actuator technology), (8)
distilled water tank, (9) smart mattress including
other non-contact sensors and electrodes, (10)
digital electronics for incubator controlling and
vital signal processing.
Note that this conceptual design comes with
an infrared thermography imaging camera embed-
ded into the incubator wall allowing monitoring
and registration of the surface temperature of the
neonate throughout the circadian rhythm. Such a
design would be possible with the advance of
miniaturized, high thermal resolution and cost-
effective infrared imagers. In addition, this design
could be accompanied by several other non-
contact physiological monitoring modalities, such
as e.g., magnetic impedance tomography (MIT)
(Heimann, 2009; Cordes, 2011), capacitive elec-
trocardiography (cECG) (Leonhardt, 2006), oxy-
gen saturation imaging (Blazek, 1993), proper
respiration monitoring and phonocardiography
acoustic imaging of newborn infant (Abbas, 2008,
2009).
Moreover, this monitoring technique will
increase the survival rate, due to the possible re-
duction of the early hypothermic shock that may

Figure 29. Futuristic concept of future neonatal incubator (a) with nursing comfortable design and in-
tegral use of imaging and sensing technologies, and (b)longitudinal section of futuristic incubator with
illustration of internal structure heat flow dynamics inside it (© MedIT, 2010)

118
Neonatal Infrared Thermography Monitoring
occur during the first hours of life. Despite the
technical difficulties in maintaining an optimal
medical thermography imaging, however, the
precision of the surface temperature will promote
the use of such measurement in various clinical
applications, such as tumors and superficial lesions
localization, peripheral neurological disorders and
ophthalmic corneal disease diagnosis.
Heat transfer dynamics optimization is also
a point of achievement for this type of futuristic
incubator design. Generally, depending on new
material, geometric profile and method of actua-
tion used for the convective incubator system,
the perspective design could avoid some of the
persisting problems of the available incubator
model in the market.
Furthermore, this new non-contact and non-
invasive respiration monitoring technique, based
on IRTR measurement, may be integrated as a
fully wireless solution for neonatal incubators.
Additionally, this method opens the closed gates
for developing a cutting-edge technology, although
the cost and technical sophistication will barrier
these efforts. However, with the advancement of
micro-infrared detector arrays and thermography
calibration optimization, this technology may have
a chance for standard use inside NICU wards.
ACKNOWLEDGMENT
We hereby express our cordial thanks to the neona-
tal intensive care unit staff, at university Hospital,
RWTH Aachen University, neonatologists and
nurses for their help, patient, and cooperation to
put this project on the track of success. Finally, we
thank the medical ethics committee at the faculty
of medicine, RWTH Aachen University for provid-
ing the acceptance to conduct the clinical study of
neonatal infrared thermography. Also we express
special thanks for Prof. Dr.-Ing Vladimir Blazek,
head of biomedical optics section in Philips chair
for medical information technology for his valuable
corrections and recommendation for this chapter.
REFERENCES
Abbas, A. K., Heimann, K., Jergus, K., Orlikowsky,
T., & Leonhardt, S. (2011). Neonatal non-contact
respiratory monitoring based on real-time infrared
thermography. Biomedical Engineering Online,
10, 93. doi:10.1186/1475-925X-10-93
Abbas, A. K., Heimann, K., Orlikowsky, T., &
Leonhardt, S. (2009). Non-contact respiratory
monitoring based on real-time IR-thermography.
In Proceedings of the World Congress of Medical
Physics and Biomedical Engineering, Munich,
Germany.
Abbas, A. K., & Leonhardt, S. (2008). NIRT
pattern classification based on ICA clustering
algorithm. In Proceedings of the Farbbildverar-
beitung Workshop, Aachen, Germany.
Adams, A. K., Nelson, R. A., Bell, E. F., & Egoavil,
C. A. (2000). Use of infrared thermographic
calorimetry to determine energy expenditure in
preterm infants. The American Journal of Clinical
Nutrition, 71(4), 969–977.
Als, H. (1998). Developmental care in the new-
born intensive care unit. Current Opinion in
Pediatrics, 10, 138–142. doi:10.1097/00008480-
199804000-00004
Arora, S. (2008). Kangaroo mother care. Nursing
Journal of India, 99(11), 248–250.
Asakura, H. (2004). Fetal and neonatal thermo-
regulation. Journal of Nippon Medical School,
71(6), 360–370. doi:10.1272/jnms.71.360
Aylott, M. (2006). The neonatal energy triangle.
Part 2: Thermoregulatory and respiratory adaption.
Paediatric Nursing, 18(7), 38–42.
Bell, E. F., Neidich, G. A., Cashore, W. J., &
Oh, W. (1979). Combined effect of radiant
warmer and phototherapy on insensible water
loss in low-birth-weight infants. The Journal of
Pediatrics, 94(5), 810–813. doi:10.1016/S0022-
3476(79)80164-X
119

Bhatia, M., Poley, J. R., Haberman, J. D., &
Boon, D. J. (1976). Abdominal thermogra-
phy in infantile and childhood liver disease.
Southern Medical Journal, 69(8), 1045–1048.
doi:10.1097/00007611-197608000-00029
Bissinger, R. L., & Annibale, D. J. (2010). Ther-
moregulation in very low-birth-weight infants
during the golden hour: results and implications.
Advances in Neonatal Care, 10(5), 230–238.
Blackburn, S., DePaul, D., Loan, L. A., Marbut, K.,
Taquino, L. T., Thomas, K. A., & Wilson, S. K. (2001).
Neonatal thermal care, part I: Survey of temperature
probe practices. Neonatal Network, 20(3), 15–18.
Blazek, F., Herdegen, L., & Stichenwirtova,
B. (1963). Indirect oximetry in the functional
diagnosis of respiratory diseases in children.
Ceskoslovenska Pediatrie, 18, 307–312.
Bohnhorst, B. (2010). Skin to skin care in the neo-
natal intensive care unit: more data regarding seri-
ously ill infants are badly needed. Commentary on
Heimann et al. Impact of skin to skin care, prone and
supine positioning on cardiorespiratory parameters
and thermoregulation in premature infants. Neo-
natology, 97(4), 318–320. doi:10.1159/000255164
Bohnhorst, B., Heyne, T., Peter, C., & Poet, S.
C. (2001). Skin-to-skin (kangaroo) care, respira-
tory control, and thermoregulation. The Jour-
nal of Pediatrics, 138, 193–197. doi:10.1067/
mpd.2001.110978
Bosque, E. M., Brady, J. P., Affonso, D. D., &
Wahlberg, V. (1995). Physiologic measures of
kangaroo versus incubator care in a tertiary-
level nursery. Journal of Obstetric, Gyneco-
logic, and Neonatal Nursing, 24(3), 219–226.
doi:10.1111/j.1552-6909.1995.tb02466.x
Browne, J. V. (2004). Early relationship envi-
ronments: physiology of skin-to-skin contact
for parents and their preterm infants. Clinics
in Perinatology, 31(2), 287–298. doi:10.1016/j.
clp.2004.04.004
120
Neonatal Infrared Thermography Monitoring
Caniou, J. (1999). Passive infrared detection:
theory and application. Boston, MA: Kluwer
Academic.
Cannon, B., & Nedergaard, J. (2004). Brown
adipose tissue: function and physiological signifi-
cance. Physiological Reviews, 84(1), 277–359.
doi:10.1152/physrev.00015.2003
Chardon, K., Cardot, V., Léké, A., Delanaud,
S., Bach, V., Dewasmes, G., & Telliez, F.
(2006). Thermoregulatory control of feeding
and sleep in premature infants. Obesity (Silver
Spring, Md.), 14(9), 1535–1542. doi:10.1038/
oby.2006.177
Christidis, I., Zotter, H., Rosegger, H., Engele,
H., Kurz, R., & Kerbl, R. (2003). Infrared ther-
mography in newborns: the first hour after birth.
Gynakologisch-Geburtshilfliche Rundschau,
43(1), 31–35. doi:10.1159/000067168
Clark, R. P., & Stothers, J. K. (1980). Neonatal
skin temperature distribution using infra-red co-
lour thermography. The Journal of Physiology,
302, 323–333.
Cloherty, J. P., Eichenwald, E. C., & Stark, A. R.
(2008). Manual of neonatal care. Riverwoods,
IL: Lippincott Williams & Wilkins.
Conlon, J. D., & Drolet, B. A. (2004). Skin
lesions in the neonate. Pediatric Clinics of
North America, 51(4), 863–888. doi:10.1016/j.
pcl.2004.03.015
Cordes, A., Santos, S. A., & Leonhardt, S. (2011).
Development of a device for measuring the sen-
sitivity area of coil arrays for magnetic induction
measurements. In Proceedings of the Conference
of the IEEE Engineering in Medicine and Biology
Society (pp. 4959-4962).
Cramer, K., Wiebe, N., Hartling, L., Crumley,
E., & Vohra, S. (2005). Heat loss prevention: a
systematic review of occlusive skin wrap for pre-
mature neonates. Journal of Perinatology, 25(12),
763–769. doi:10.1038/sj.jp.7211392
Neonatal Infrared Thermography Monitoring
de Curtis, M. D., Calzolari, F., Marciano, A.,
Cardilli, V., & Barba, G. (2008). Comparison
between rectal and infrared skin temperature in the
newborn. Archives of Disease in Childhood. Fetal
and Neonatal Edition, 93(1), 55–57. doi:10.1136/
adc.2006.114314
Derflinger, T., & Kane, R. (2009). Thermal im-
aging for medical applications. Retrieved from
http://www.infraredcamerasinc.com/medical-
thermal-imaging.html
Dodd, S. R., Lancaster, G. A., Craig, J. V.,
Smyth, R. L., & Williamson, P. R. (2006). In a
systematic review, infrared ear thermometry for
fever diagnosis in children finds poor sensitivity.
Journal of Clinical Epidemiology, 59(4), 354–357.
doi:10.1016/j.jclinepi.2005.10.004
Dzukou, T., de la Pintière, A., Bétrémieux, P., Vittu,
G., Roussey, M., & Tietche, F. (2004). Kangaroo
mother care: bibliographical review on the current
attitudes, their interests and their limits. Archives
of Pediatrics, 11(9), 1095–1100. doi:10.1016/j.
arcped.2004.01.032
Eggert, P., Stick, C., & Schröder, H. (1984). On
the distribution of irradiation intensity in photo-
therapy. Measurements of effective irradiance in
an incubator. European Journal of Pediatrics,
142(1), 58–61. doi:10.1007/BF00442593
Fanaroff, A. A., Stoll, B. J., Wright, L. L., Carlo,
W. A., Ehrenkranz, R. A., & Stark, A. R. (2007).
Trends in neonatal morbidity and mortality for
very low birth weight infants. American Journal
of Obstetrics and Gynecology, 196, 147–148.
doi:10.1016/j.ajog.2006.09.014
Fei, J., & Pavlidis, I. (2006). Analysis of breathing
air flow patterns in thermal imaging. In Proceed-
ings of the Conference of the IEEE Engineering in
Medicine and Biology Society (Vol. 1, pp. 946-952).
Fei, J., & Pavlidis, I. (2007). Virtual thermistor. In
Proceedings of the Conference of the IEEE Engineer-
ing in Medicine and Biology Society (pp. 250-253).
Flenady, V. J., & Woodgate, P. G. (2003).
Radiant warmers versus incubators for regu-
lating body temperature in newborn infants.
Cochrane Database of Systematic Reviews,
(4): CD000435.
Fortuna, E. L., Carney, M. M., Macy, M., Stanley,
R. M., Younger, J. G., & Bradin, S. A. (2010). Ac-
curacy of non-contact infrared thermometry versus
rectal thermometry in young children evaluated
in the emergency department for fever. Journal
of Emergency Nursing: JEN, 36(2), 101–104.
doi:10.1016/j.jen.2009.07.017
Fransson, A.-L., Karlsson, H., & Nilsson, K.
(2005). Temperature variation in newborn babies:
importance of physical contact with the mother.
Archives of Disease in Childhood. Fetal and
Neonatal Edition, 90(6), 500–504. doi:10.1136/
adc.2004.066589
Garbey, M., Sun, N., Merla, A., & Pavlidis, I.
(2007). Contact-free measurement of cardiac pulse
based on the analysis of thermal imagery. IEEE
Transactions on Bio-Medical Engineering, 54(8),
1418–1426. doi:10.1109/TBME.2007.891930
Gaussorgues, G. (1994). Infrared thermography
(pp. 101–124). Heidelberg, Germany: Springer
Verlag. doi:10.1007/978-94-011-0711-2
Gencer, N. G., Ider, Y. Z., & Williamson, S.
J. (1996). Electrical impedance tomography:
induced-current imaging achieved with a
multiple coil system. IEEE Transactions on
Bio-Medical Engineering, 43(2), 139–149.
doi:10.1109/10.481983
Gerber, A. U. (2006). Clinical thermometry-
theory and critical approach. Therapeutische
Umschau, 63(10), 627–632. doi:10.1024/0040-
5930.63.10.627
Gibbons, C., Geller, S., & Glatz, E. (1998). Bio-
medical equipment in the neonatal intensive care
unit: Is it a stressor? The Journal of Perinatal &
Neonatal Nursing, 12, 67–73.
121

Ginalski, M. K., Nowak, A. J., & Wrobel,
L. C. (2007). A combined study of heat and
mass transfer in an infant incubator with an
overhead screen. Medical Engineering &
Physics, 29(5), 531–541. doi:10.1016/j.medeng-
phy.2006.07.011
Ginalski, M. K., Nowak, A. J., & Wrobel, L.
C. (2008). Modeling of heat and mass trans-
fer processes in neonatology. Biomedical
Materials (Bristol, England), 3(3), 034113.
doi:10.1088/1748-6041/3/3/034113
Hanssler, L., & Breukmann, H. (1992).
Measuring skin temperature in premature
infants: Comparison of infrared tele-ther-
mography and electric contact thermom-
etry. Klinische Padiatrie, 204(5), 355–358.
doi:10.1055/s-2007-1025371
Heimann, K., Steffen, M., Bernstein, N., Heerich,
N., Stanzel, S., & Cordes, A. (2009). Non-contact
monitoring of heart and lung activity using mag-
netic induction measurement in a neonatal animal
model. Biomedical Technology, 54(6), 337–345.
doi:10.1515/BMT.2009.044
Heny, C., & Frizet, M. (2003). Design consider-
ations for a medical thermographic expert system.
In Proceedings of the 25th Annual International
Conference of the IEEE Engineering in Medicine
and Biology Society, Cancun, Mexico.
Horns, K. M. (2003). Innovative non-invasive
measures of thermal stability in low birth weight
infant (VLBW). In Proceedings of the InfraMa-
tion Conference.
Huang, K. N., Huang, C. E., Li, Y. C., & Young,
M. S. (1999). A new method for temperature Mea-
surement in an Infant Incubator. In Proceedings of
the Biomedical Engineering Society Conference,
Atlanta, GA.
Hyman, W. A. (2006). Infant incubators. Biomedi-
cal Instrumentation & Technology, 40(4), 254.
doi:10.2345/i0899-8205-40-4-254.1
122
Neonatal Infrared Thermography Monitoring
Ibe, O. E., Austin, T., Sullivan, K., Fabanwo,
O., Disu, E, & del Costello, A. M. (2004). A
comparison of kangaroo mother care and con-
ventional incubator care for thermal regulation
of infants < 2000 g in Nigeria using continu-
ous ambulatory temperature monitoring. An-
nals of Tropical Paediatrics, 24(3), 245–251.
doi:10.1179/027249304225019082
Jamsa, K., & Jamsa, T. (1998). Technology in
neonatal intensive care —A study of parents’
experiences. Technology and Health Care, 6,
225–230.
Johnson, K. J., Bhatia, P., & Bell, E. F. (1991).
Infrared thermometry of newborn infants. Pedi-
atrics, 87(1), 34–38.
Kent, A. L., & Williams, J. (2008). Increasing am-
bient operating theatre temperature and wrapping
in polyethylene improves admission temperature
in premature infants. Journal of Paediatrics and
Child Health, 44(6), 325–331. doi:10.1111/j.1440-
1754.2007.01264.x
Knobel, R., & Holditch-Davis, D. (2010). Thermo-
regulation and heat loss prevention after birth and
during neonatal intensive-care unit stabilization
of extremely low-birth weight infants. Advances
in Neonatal Care, 10(5), 7–14.
Kumar, A., & Bhat, B. V. (1996). Respiratory
distress in newborn. Indian Journal of Maternal
and Child Health, 7(1), 8–10.
Leonhardt, S., Aleksandrowicz, A., & Steffen, M.
(2006). Magnetic and capacitive monitoring of
heart and lung activity as an example for personal
healthcare. In Proceedings of the 3rd IEEE/EMBS
International Summer School on Medical Devices
and Biosensors (pp. 57-60).
Libert, J. P., Bach, V., & Farges, G. (1997). Neutral
temperature range in incubators: performance of
equipment in current use and new developments.
Critical Reviews in Biomedical Engineering,
25(4-5), 287–370.
Neonatal Infrared Thermography Monitoring
McBride, T. O., Pogue, B. W., Gerety, E. D., Poplack,
S. B., Osterberg, U. L., & Paulsen, K. D. (1999).
Spectroscopic diffuse optical tomography for the
quantitative assessment of hemoglobin concentration
and oxygen saturation in breast tissue. Applied Op-
tics, 38(25), 5480–5490. doi:10.1364/AO.38.005480
McIntosh, J. R. (2005). Ancient Mesopotamia:
New perspectives, understanding ancient civiliza-
tion. Santa Barbara, CA: ABC-CLIO.
Merenstein, G. B., & Gardener, S. L. (1998).
Handbook of neonatal intensive care (4th ed.).
St. Louis, MO: Mosby.
Miles, M., Funk, S. G., & Kasper, M. A. (1991).
The neonatal intensive care unit environment:
Sources of stress for parents. American Associa-
tion of Critical Care Nurses, Clinical Issues, 2(2),
346–354.
Moseley, H., & Marshall, J. (1988). Protec-
tive light shields for neonatal phototherapy.
Lancet, 2(8615), 854–855. doi:10.1016/S0140-
6736(88)92821-8
Museux, N., Cardot, V., Bach, V., Delanaud,
S., Degrugilliers, L., & Agourram, B. (2008). A
reproducible means of assessing the metabolic
heat status of preterm neonates. Medical Physics,
35(1), 89–100. doi:10.1118/1.2815966
Nimah, M. M., Bshesh, K., Callahan, J. D., &
Jacobs, B. R. (2006). Infrared tympanic ther-
mometry in comparison with other temperature
measurement techniques in febrile children.
Pediatric Critical Care Medicine, 7(1), 48–55.
doi:10.1097/01.PCC.0000185476.35550.B2
Owa, J. A., Adebami, O. J., Fadero, F. F., & Slusher,
T. M. (2011). Irradiance readings of phototherapy
equipment: Nigeria. Indian Journal of Pediatrics,
78(8), 996–998. doi:10.1007/s12098-011-0382-4
Pavlidis, I., & Levine, J. (2002). Thermal image
analysis for polygraph testing. IEEE Engineering
in Medicine and Biology Magazine, 21(6), 56–64.
doi:10.1109/MEMB.2002.1175139
Pollina, D. A., Dollins, A. B., Senter, S.
M., Brown, T. E., Pavlidis, I., Levine, J. A.,
& Ryan, A. H. (2006). Facial skin surface
temperature changes during a “concealed
information” test. Annals of Biomedical
Engineering, 34(7), 1182–1189. doi:10.1007/
s10439-006-9143-3
Rojas, R. D. (2006, October). Dynamic estimation
of core temperature in premmies using an adaptive
observer. In Proceedings of the First Joint BMES/
EMBS Conference (p. 1012).
Saxena, A. K., & Willital, G. H. (2008). Infra-
red thermography: experience from a decade of
pediatric imaging. European Journal of Pedi-
atrics, 167(7), 757–764. doi:10.1007/s00431-
007-0583-z
Silverman, W. A. (2007). Internet online articles
on neonatology medicine. Retrieved from http://
www.neonatology.org/classics/silverman/silver-
man1.html
Sun, N., Pavlidis, I., Garbey, M., & Fei, J. (2006).
Harvesting the thermal cardiac pulse signal.
Medical Image Computing and Computer Assisted
Intervention, 9(2), 569–576.
Teich, J. S. (1996, November). Digital infra-
red imaging for medicine recent advances in
IR focal plane array imaging technology. In
Proceedings of the 18th Annual International
Conference of the IEEE Engineering in Medi-
cine and Biology Society, Amsterdam, The
Netherlands.
Vollmer, M., & Möllmann, K.-P. (2010). Infrared
thermal imaging: Fundamentals, research and
applications. Berlin, Germany: Wiley-VCH.
doi:10.1002/9783527630868
Zhu, Z., Tsiamyrtzis, P., & Pavlidis, I. (2007).
Forehead thermal signature extraction in lie
detection. In Proceedings of the Conference of
the IEEE Engineering in Medicine and Biology
Society (pp. 243-246).
123

ADDITIONAL READING
Bronzino, J. D. (2006). Medical devices and sys-
tem. Boca Raton, FL: CRC/Taylor and Francis.
Dereniak, E. L., & Boreman, G. D. (1996). Infra-
red detectors and systems. Ann Arbor, MI: The
University of Michigan.
Derflinger, T. The Thermogram Center, & Kane, R.
(2009). Thermal imaging for medical applications.
Retrieved from http://www.infraredcamerasinc.
com/medical-thermal-imaging.html
Diakides, N. A., & Bronzino, J. D. (2007). Medical
infrared imaging. Boca Raton, FL: CRC Press.
doi:10.1201/9781420008340
Hey, E. N., & Katz, G. (1970). The optimum
thermal environment for naked babies. Archives
of Disease in Childhood, 45, 328. doi:10.1136/
adc.45.241.328
124
View publication stats
Neonatal Infrared Thermography Monitoring
Jacques, P. A., & Viccari, R. M. (2006). Consid-
ering students’ emotions in computer-mediated
learning environments. In Ma, Z. (Ed.), Web-based
intelligent e-learning systems: Technologies and
applications (pp. 122–138). Hershey, PA: Infor-
mation Science.
Kaplan, H. (2007). Practical applications
of infrared thermal sensing and imaging
equipment. Bellingham, WA: SPIE Press.
doi:10.1117/3.725072
Maldague, X. (2001). Theory and practice of
infrared technology for nondestructive testing.
New York, NY: John Wiley & Sons.
Weiss, M. E., Poeltler, D., & Gocka, I. (1994).
Infrared tympanic thermometry for neonatal tem-
perature assessment. Journal of Obstetric, Gyne-
cologic, and Neonatal Nursing, 23(9), 798–804.
doi:10.1111/j.1552-6909.1994.tb01955.x