Content uploaded by Domenico Formica
Author content
All content in this area was uploaded by Domenico Formica on May 25, 2022
Content may be subject to copyright.
IEEE SENSORS JOURNAL, VOL. 13, NO. 11, NOVEMBER 2013 4561
Ecological Sucking Monitoring of Newborns
Fabrizio Taffoni, Member, IEEE, Eleonora Tamilia, Member, IEEE, Maria Rosaria Palminteri,
Emiliano Schena, Member, IEEE, Domenico Formica, Member, IEEE, Jonathan Delafield-Butt,
Flavio Keller, Sergio Silvestri, Member, IEEE, and Eugenio Guglielmelli, Senior Member, IEEE
Abstract— Feeding by sucking is one of the first activities of
daily life performed by infants. Sucking plays a fundamental
role in neurological development and may be considered a good
early predictor of neuromotor development. In this paper, a new
method for ecological assessment of infants’ nutritive sucking
behavior is presented and experimentally validated. Preliminary
data on healthy newborn subjects are first acquired to define
the main technical specifications of a novel instrumented device.
This device is designed to be easily integrated in a commercially
available feeding bottle, allowing clinical method development
for screening large numbers of subjects. The new approach
proposed allows: 1) accurate measurement of intra-oral pressure
for neuromotor control analysis and 2) estimation of milk volume
delivered to the mouth within <2% variation between estimated
and reference volumes.
Index Terms—Instrumented objects, suckling monitoring,
ecological assessment, nutritive sucking, motor control.
I. INTRODUCTION
NEUROPHYSIOLOGICAL development in infants may
be indirectly assessed by the study of movement.
Precthl demonstrated spontaneous motility of infants could
be regarded as the expression of spontaneous neural activity,
presenting an excellent marker of neural dysfunctions [1].
Early motor acts and their development provide the infant
with new experiences and opportunities for exploration of the
world, and world exploration in turn creates context for the
development of the brain [2]. This process is critical, since
the progressive acquisition of motor skills provides infants
with an increasingly growing set of opportunities for acquiring,
practicing and refining abilities, that is an essential prerequisite
Manuscript received March 5, 2013; revised June 17, 2013; accepted
June 22, 2013. Date of publication June 27, 2013; date of current version
October 2, 2013. This work was supported by the Italian Ministry of
Education, Universities and Research under the Futuro in Ricerca Research
Program TOUM project under Grant B81J10000160008. The associate editor
coordinating the review of this paper and approving it for publication was
Dr. Anupama Kaul.
F. Taffoni, E. Tamilia, M. R. Palminteri, D. Formica, and E. Guglielmelli
are with the Laboratory of Biomedical Robotics and Biomicrosystems,
Università Campus Bio-Medico di Roma, Rome 00128, Italy (e-mail:
f.taffoni@unicampus.it; e.tamilia@unicampus.it; m.palminteri@unicampus.it;
d.formica@unicampus.it; e.guglielmelli@unicampus.it).
E. Schena and S. Silvestri are with the Center for Integrated Research,
Unit of Measurements and Biomedical Instrumentation, Università Campus
Bio-Medico di Roma, Rome 00128, Italy (e-mail: e.schena@unicampus.it;
s.silvestri@unicampus.it).
J. Delafield-Butt is with the Perception Movement Action Consortium,
University of Edinburgh, Edinburgh EH8 8AQ, U.K., and also with the
University of Strathclyde, Glasgow G4 0LT, U.K. (e-mail: jonathan.delafield-
butt@strath.ac.uk).
F. Keller is with the Laboratory of Developmental Neuroscience and
Neural Plasticity, Università Campus Bio-Medico di Roma, Rome 00128, Italy
(e-mail: f.keller@unicampus.it).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/JSEN.2013.2271585
for growth in other domains, especially those of language
and social interaction [3]. Therefore, the investigation of early
motor skills in newborns provides a window not only on
the development of motor control, but more generally on the
neural dynamics that underpin neuropsychological health for
development and learning [4].
Instruments and methods typically used to investigate learn-
ing and also re-learning capabilities in human adult subjects
[5]–[7] are not suitable for newborn infants because of their
obtrusiveness and complex equipment. Recently developed
instrumented toys and other objects, possibly embedding
purposively developed microsensors [8], [9], offer a great
possibility for ecological non-obtrusive investigation of brain
development in infancy and childhood [10]–[14]. Despite cur-
rent researches on novel design methodologies for advanced
wearable systems [15], such technologies are not suitable
for newborn infants with their comparatively poor motor
repertoires and relative disinterest in manipulating objects.
Sucking, on the other hand, is one of the first and most
frequent action performed by infants [16], [17]. It begins in
uterus from gestational age 15 to 18 weeks [18] and usually
becomes stable and well patterned by 34 weeks in preparation
for birth [19].
Nutritive sucking (NS) postpartum allows an infant to
obtain food. It consists of a series of bursts and pauses; each
burst contains several suck cycles that occur at approximately
1 Hz [20]. A nutritive suck is characterized by the alternation
of suction and expression [21]: suction defines the negative
intraoral pressure generated by the infant in order to draw milk
into the mouth and expression corresponds to the compression
and/or stripping of the nipple between the tongue and the hard
palate as milk is ejected into the mouth [22], [23].
The act of sucking is a complex task for the neonate,
because it requires not only an efficient sucking ability, but
also a proper coordination between the rhythmic processes of
sucking, swallowing and respiration [24]–[26]. This coordina-
tion requires intact brainstem pathways and transmission of
impulses through the cranial nerves to healthy musculature in
the mouth, tongue, and pharynx [27]. Lack of coordination
explains changes in the sucking rate and the appearance of
abnormal clinical signs such as low consumption of food,
choking, regurgitation, vomiting or respiratory disorders. For
this reason, the assessment of the sucking pattern represents
one of the most accessible and earliest possible cues of
infant neurophysiological status, as shown for the first time
by Wolff [20], followed by several others [26]–[29]. Precise
analyses of sucking can provide not only valuable insights
into the integrity of an infant’s central nervous system, but
can also serve as an early prognostic indicator of future
1530-437X © 2013 IEEE
4562 IEEE SENSORS JOURNAL, VOL. 13, NO. 11, NOVEMBER 2013
neurodevelopmental disorder. Mizuno et al. [27] found that
patterns and values of expression and sucking pressures during
neonatal feeding can predict neurodevelopmental outcomes
evaluated by the Bayley Scales of Infant Development (BSID)
at 18 months. Medoff-Cooper [30] as well reported a signifi-
cant relationship between sucking organization (characterized
only by sucking pressure values) and BSID values, evaluated
at the earlier age of 12 months. Craig et al. [31] reported
evidence that the degree of control of the increasing and
decreasing suction pressures generated near corrected term in a
prematurely-born ‘at risk’ cohort predicted motor development
measured at 6 months using postural, fine, and gross motor
assessment.
These studies stimulate further development of protocols
for neonatal sucking assessment as an indicator of infants’
neurophysiological health or dysfunction, and may help to
address an unmet need for a simple, cost-effective screening
strategy for early identification of infants at risk for cognitive
and psychomotor delays [32].
Hence, driven by this need, the aim of this study has been
to develop a new methodology for an objective measurement
of sucking parameters and to integrate such a methodology in
a portable, non-intrusive, low-cost device.
II. FUNCTIONAL AND TECHNICAL SPECIFICATIONS
Analysis of intra-oral pressure changes during newborn NS
may represent one of the earliest easily accessible assessments
of infant neurophysiological status [30]. Further, simultaneous
assessment of the volume of ingested milk may allow quantifi-
cation of the NS efficiency for improved assessment [33]. Yet,
the methodological solutions previously adopted, i.e.in the
NS studies cited above, were based on complex, non-portable
technologies instrumented with ultrasound systems, physio-
logical pressure monitoring, or with additional equipment for
continuous measurement by weight of the milk reservoir.
These systems previously employed were complex, expensive,
and awkward to deploy. They remain restricted to small-scale
research studies and do not allow for extended application
to large-scale studies or deployment to everyday domestic or
clinical settings. Such deployment is desirable for method-
ological and analytical improvements that may ultimately
contribute significant new data to clinical assessment of infant
neurophysiological health. Two principal modalities of data on
infant sucking are apparent: (i) continuous monitoring of intra-
oral pressure and (ii) continuous monitoring of expressed milk
volume. For these reasons the proposed device should monitor
intra-oral pressure and volume of taken milk in an ecological,
unobtrusive way, even at home during regular bottle feeding.
For ecological we mean a methodology and/or an instrumen-
tation that don’t modify the environment and/or the observed
task. In the feeding task, for example, the inclination of the
bottle with respect to gravity is particularly important because
low inclination may not completely fill the nipple with milk,
while, in case of high inclination (>60°), the infant may expe-
rience high hydrostatic pressure, especially at the beginning
of feeding. Emphasis on ecological nature of data collection
is common in behavioral studies and becomes particularly
Fig. 1. Intra-oral pressure of a 1 week newborn recorded during a feeding
task.
important for infants who may already be distressed and can
not cooperate with experimenters or clinicians with additional,
complex and time-consuming settings and procedures.
For example, another possibility for measuring the milk
volume ingested is to use some simple non-technological
approaches, such as weighing the bottle before and after
the feed or checking the volume of ingested milk by visual
inspection of the scale on the side of the bottle. The main
drawback of these solutions is that they allow a global esti-
mation of the ingested volume, but they do not continuously
assess milk volume intake; therefore, they do not allow an
energetic analysis of sucking. Simultaneous recording of both
intraoral pressure and milk volume intake overcomes this
limitation [34]. Moreover, the assessment of the rate of milk
flow during bottle feeding it is known to play a crucial role for
clinical evaluation: it is one of the most important determinants
of feeding-related ventilator changes in some term and preterm
infants, where the efficient feeding and breathing assistance is
crucial for physiological and development health [35].
In order to define the specifications of the device that
should simultaneously and continuously measure intra-oral
pressure and milk volume intake preliminary data analysis was
undertaken to verify intra-oral pressure range and bandwidth.
Eight term-birth, healthy infants were recorded while feeding
shortly after birth (mean age 2.6 days, std. dev. 1.5 days) and
four were recorded approximately 8 weeks later (mean age
55 days, std. dev. 15 days). Intra-oral pressure was measured
by inserting an umbilical catheter 40 cm in length with an
internal diameter of 1.0 mm and external diameter of 1.7 mm
(Ref. 1270.05, Vygon, France) through the teat (Standard Teat,
Cow & Gate, UK) of a feeding bottle so that it protruded
approximately 2 mm from its tip into the middle of the infant’s
oral cavity. The catheter was primed with distilled water
and connected to a pressure transducer (TranStar, Medex,
UK/France/Italy/Germany) and medical physiology monitor
(Datex-Ohmeda Light, Finland). Data were captured to a lap-
top PC using proprietary software (Collect4, General Electric
Healthcare, Finland).
Fig. 1 reports a typical 10 seconds pressure burst: experi-
mental data confirm that intra-oral pressure is in the range of
TAFFONI et al.: ECOLOGICAL SUCKING MONITORING OF NEWBORNS 4563
Fig. 2. Power Spectral Density of intra-oral pressure during sucking. In the
example reported above the PSD is below 20 Hz. The peak below 5 Hz is due
to Nutritive Sucking (NS) whereas the peak at about 6 Hz is due to breathing.
[−140 +15] mmHg reported in literature [36]. The bandwidth
of the pressure signal has been estimated calculating its Power
Spectral Density (PSD) by means of the Welch overlapped
segmented average. As shown by Fig. 2 the bandwidth may
be considered well below 20 Hz, which has been taken as
the required bandwidth of our device. The volume of milk
ingested by an infant during each suck or burst may depend on
several factors, e.g., the infant’s age, weight, postconceptional
age (PCA) at birth.
Table I reports the values of several sucking parame-
ters obtained by several authors in their studies [24], [33],
[36]–[38]. The values reported by Taki et al. in their recent
study [36] have been considered in the present work because
they report both the volume of a single suck and the volume
of an entire burst, unlike Fadavi’s [37] and Lang’s [38]
works, and because they also report on longitudinal sucking
performance, allowing for consideration of differences from
1 to 6 months of age.
According to [36], a bottle-fed infant ingests
0.28 ±0.13 mL per suck at the postnatal age of 1 month. This
value slightly increases with the age, reaching 0.30 ±0.24 mL
at 3 months and 0.40 ±0.20 mL at 6 months. These values,
together with the number of sucks per burst, allow estimation
of the minimum volume of milk an infant can ingest in a
suck or a burst. During the first 6 months of life the average
value of 0.2 mL has been considered a minimum volume per
suck, whereas 8 mL is considered the minimum volume per
burst.
III. DESIGN AND FABRICATION OF SUCTION DEVICE
A. General Architecture
The two main functional requirements for the sucking
device are its portability and ease of use. For this reason, a
smart module was designed and developed to integrate into a
commercial infant feeding bottle. Fig. 3 reports the general
architecture of the module. A key idea was to physically
decouple the sensing electrical part from the milk reservoir.
To this end the electronics were separated from the milk
reservoir by a semipermeable membrane (this allows air flow,
but blocks the liquid) placed between the reservoir and the
TAB L E I
VALUES OF PARAMETERS RELATED TO SUCKING:REVIEW OF THE MAIN
STUDIES
Fig. 3. The general architecture of the device. Electronic components are
separated from the milk reservoir thanks to a membrane. A Mechanical Safety
Device (MSD) is added to the milk reservoir to avoid the creation of excessive
negative pressures inside the bottle higher beyond a threshold level.
electronics: direct contact between the milk and the electronics
was avoided, solving both important problems of contamina-
tion and electrical safety.
The sensing core had to integrate sensors allowing measure-
ment of intra-oral pressure and volume of ingested milk. The
components of the sensing core will be discussed in detail in
Section III-B below. Data coming from sensing core needed
to be acquired by the control unit and stored on a suitable
support. To allow portability, robustness and guarantee long
duration monitoring, a micro SD support was chosen. For this
reason the control unit was required have the RAM necessary
for writing/reading operations on a micro SD: for FAT16
micro-SD, 512 MB RAM memory was deemed necessary and
a PIC18F46J50 chosen to be used.
Finally, a mechanical safety device was provided on the
bottom of the bottle in order to avoid conditions that may
hinder the infant’s sucking. This is illustrated in Section III-C
below.
4564 IEEE SENSORS JOURNAL, VOL. 13, NO. 11, NOVEMBER 2013
B. Sensing Core
The intraoral sucking pressure is measured with a modi-
fied disposable teat connected via a non-compliant catheter
(1.2 mm of internal diameter) to a low cost integrated silicon
pressure sensor (MPX7025, Freescale Semiconductor). This
sensor has a range of ±25 kPa (about ±190 mmHg), well
above the physiological range of intraoral pressures (Fig. 1).
The catheter was inserted into the teat through an incision
at its base. The tube was then threaded into the inside of the
teat and out through one of the three manufactured feeding
holes at the top. To be able to measure the intraoral sucking
pressure, the tube was positioned so that it protruded about
2 mm through the end of the teat as reported in [37].
The estimation of the volume of milk delivered to the infant
during each sucking act is particularly important because it
allows quantification of the effectiveness of each NS.
It is very common in pediatric practice to weigh the bottle
at the beginning and at the end of the feeding to obtain a
raw estimation of the overall amount of milk delivered to the
child. In the above cited researches, authors provided complex
measurement systems to refine this technique. The proposed
systems are quite complex and expensive. In this work we
propose a low cost method to estimate the volume of milk
delivered to the newborn based on a measurement of pressure.
During a single suck, an infant closes her/his lips around
the bottle teat and, as the jaw and tongue drop down, produces
a negative pressure inside the oral cavity that causes the milk
flow towards the mouth. In this phase, as long as a good seal
around the teat is maintained, no air can flow inside the bottle.
The flow of milk toward the mouth causes the generation
of a gradual negative pressure inside the bottle which persists
until the newborn releases her/his lips allowing air to flow
inside the bottle. Therefore, by measuring the pressure of a gas
(the air inside the bottle), an estimation of the milk volume
delivered to the infant during each suck may be obtained, as
reported in other medical fields [39].
In general, the relationship between volume and pressure of
a gas may be expressed as a function of number of moles (n),
gas volume (V) and temperature (T):
P=P(n,V,T)(1)
Taking into account all the different variables, the pressure
drop (dP) within the bottle can be expressed as:
dP =∂P
∂ndn +∂P
∂VdV +∂P
∂TdT (2)
During each suck, the number of moles of the gas inside the
bottle may be considered constant because there is not air
exchange between the bottle and the environment (dn =0).
Moreover, variation of air temperature may be hypothesized
as negligible during a suck. Taking into account these assump-
tions, and considering the air inside the bottle as an ideal gas
(PV =nRT), equation (2) may be rewritten to estimate the
variation of air volume (dV) inside the bottle as:
dV ∼
=−V2
nRT dP (3)
Fig. 4. Feeding bottle: at time tithere is an inner air volume equal to Vi;
after one suck, a volume variation equal to dViwill be obtained.
V in this equation represents the volume of air inside the bottle
before the suck. This volume may be updated using a recur-
sive algorithm that increments its value by the estimate dV.
Referring to Fig. 4, it will be:
V1=V0+dV0
V2=V1+dV1
...
Vi+1=Vi+dVi(4)
where V0is the initial volume of air inside the bottle.
The change in volume of air for each suck will be equal
(disregarding the sign) to the volume of milk delivered to the
child.
The term nRT in absence of air flow may be considered
constant and equal to the product between P and V (according
to the Boyle-Mariotte’s law), for this reason (3) may be
expressed as:
dV =V
PdP (5)
V0may be estimated producing a perturbation of known
volume (V) and measuring the air pressure before (P1)and
after (P2) the volume variation. It will be:
P1V0=A
P2(V0+V)=A(6)
From the previous ones, the unknown V0will be:
V0=P2V
P1−P2(7)
in this way a semiautomatic initialization of the device will be
possible enabling its use ecologically in non-structured envi-
ronments. In order to estimate the volume of milk delivered to
the infant during the sucking acts, using the method described
above, the sensing core has been equipped with a second
pressure sensor, MPX7025.
C. Integration
The sensing core has been integrated on a commercially
available infant feeding bottle. For solving the contamination
and electrical safety problems, described in section III-A, a
commercial bottle was used (First Bottle, MAM Babyartikel
GesmbH, Vienna, Austria) with a vented base with eight air
holes and an embedded silicone membrane that blocks liquid
TAFFONI et al.: ECOLOGICAL SUCKING MONITORING OF NEWBORNS 4565
Fig. 5. Instrumented feeding bottle: A) prototype; B) 3D-CAD of the
prototype. In B it is possible to observe: 1 external electronic support, 2
silicon membrane; 3 MSD; 4 electronics.
but not air, allowing an actual separation of the electronic
smart part from the milk reservoir. The vented base was experi-
mentally modified to prevent air entering into the bottle as milk
flowed out, to satisfy the method described in section III-B.
Hence, all the holes except one (the one that will be needed for
the intra-bottle pressure measurement) were sealed with a non-
toxic smooth consistency mouldable silicone rubber (RTV-530,
Prochima S.r.l.). Due to this modification, if the newborn does
not release his/her lips around the teat, a decreasing pressure
inside the bottle will generate. If such pressure reaches the
equilibrium with the opposing suction pressure exerted by the
infant, milk flow may be hampered [40].
Since such conditions would alter the infant’s sucking
behavior, an additional mechanical safety device has been
included at the bottle base, with the aim of preserving the
newborn from excessive fatigue rather than estimating the
volume of milk taken in under such altered conditions. The
device is a polypropylene check valve (Check Valve PP695-
2B2BF, Coast Pneumatics Inc.) with a 60 mmHg cracking
pressure; it allows the air to flow inside the bottle when the
negative pressure inside it exceeds the value of −60 mmHg.
A series of tests have been carried out on the valve, showing
that 1 s after its opening the mean intra-bottle pressure increase
is 10.7 ±0.4 mmHg, preventing in this way the negative
pressure inside the bottle to rise above values that can cause
excessive fatigue to the infant, in case she/he does not release
the teat.
To allow for bottle sterilization in between uses, an external
support for the electronics has been designed. This support
consists of two parts connected through a bayonet closure:
one part is attached to the bottle base, whereas the other one,
including the electronics, can be easily removed (see Fig. 5)
allowing for easy washing and sterilization of the bottle.
IV. EXPERIMENTAL VALIDATION
Experimental validation of the method for volume esti-
mation proposed in section III-B has been carried out as
described below. The experimental setup used is shown in
Fig. 6. A graduated reservoir was filled with water at room
temperature. The reservoir was connected to the atmosphere
by means of a manual on/off valve. The pressure sensor
Fig. 6. Experimental setup used for empirical validation of volume estimation
method presented in section III.B.
Fig. 7. Pressure variation for different air volumes inside the bottle. The
subtracted liquid volume is equal to 0.2 mL. The dotted red line represents
the power regression model.
MPXV7025 is used to measure the air pressure. The base of
the reservoir is connected by means of three-way valve to
a syringe (Vmax =1 mL, resolution 0.01 mL). The manual
on/off valve simulates the child’s lips: when the lips are closed
around the teat the valve is closed, when the lips are released,
allowing air flow inside the bottle, the valve is opened. The
syringe is used to simulate suck and burst allowing intake from
0.1 mL to 1 mL of liquid.
According to (5) and for a fixed V, the measured pressure
variation is inversely proportional to the inner air volume.
For this reason, the maximal inner air volume compatible
with pressure sensor resolution (about 1 mmHg) needed to
be estimated. A fixed volume of water equal to 1 suck
(0.2 mL) was subtracted for ten times to the reservoir and the
corresponding pressure drop measured. After each subtraction,
the water was reinserted into the reservoir and the on/off
valve opened to allow atmospheric pressure in the chamber.
Trials have been repeated at different initial air volumes:
25, 35, 45, 55, 65 mL, to simulate the reduction of liquid
inside the bottle due to nutrition. Results of these trials are
reported in Fig. 7: as reported in (5), the pressure drop was
inversely proportional to the inner air volume and may be
fitted with a power law ( f(x)=ax∧b,a=167.3, b=
−1.034, R2=0.99). According to this fit, the maximal inner
volume compatible with pressure sensor resolution was equal
to 140 mL, corresponding to a volume of ingested milk equal
to 115 mL (almost two times higher than volume of milk
ingested by healthy newborns).
4566 IEEE SENSORS JOURNAL, VOL. 13, NO. 11, NOVEMBER 2013
TAB L E II
VESTIMATION AT DIFFERENT INNER AIR VOLUMES
Fig. 8. Volume variation estimation for different air volume inside the bottle.
The dotted red line represents the actual liquid volume variation, equal to
0.2 mL. The dash-dotted black lines represent the actual value ±2%.
In order to assess the performances of the proposed method,
a set of measurements at the same inner air volumes reported
above was carried out, for a total variation of liquid equal
to 40 mL. Ten preliminary measurements with 25 mL of air
were performed to estimate the initial V0 according to (7).
The estimated volume was 25.59 ±0.07 mL (percent error =
2.4%). For each inner air volume 0.2 mL of liquid (equal to
1 suck) was subtracted by the reservoir (see Fig. 6.A), then
the reservoir was closed using the three way valve and the
syringe emptied in an external tub (see Fig. 6.B). After 15
sucks the on/off valve was opened simulating the releasing of
the lips around the teat and the consequent air flow allowed
inside the bottle. For each suck, the volume of subtracted
liquid was estimated according to (3) and compared with the
actual subtracted volume. Trials were repeated at different
inner air volumes to verify if the reduction of sensitivity
may affect the estimation of volume in the considered range
(see Table II).
In Fig. 8 volume estimation results (blue points) are com-
pared with the actual volume subtracted (red dotted line).
Despite decrease of sensitivity with increasing inner air vol-
ume, the error remains small and does not exceed the 2% of
the actual volume subtracted.
V. CONCLUSION
The act of sucking is a complex task for newborns, requiring
necessary strength of all oropharyngeal structures involved as
well as adequate maturation of the relevant motor circuits
within the central nervous system. For this reason it represents
a good candidate for indirect assessment of neurophysiological
development of newborns.
In this work, a new methodological advance for newborn
sucking monitoring has been presented and experimentally
validated. A low-cost electronic has been designed and inte-
grated into a commercially available bottle. The prototype
instrument allows continuous measurement of intraoral pres-
sure and continuous estimation of volume of milk delivered to
the newborn, to within a 2% margin of error, during feeding.
Volume estimation is based on measurement of the air pres-
sure inside the feeding bottle. Moreover, volume estimations
are independent from the tilt of the bottle with respect to
gravity, so that it is not needed any orientation monitoring,
or keeping the bottle at a specific angle to obtain good
volume estimations, as done by [41]. Since the use of our
new device does not modify how bottle feeding is commonly
done (e.g. by altering the tilting angle) and since it does not
need complex calibration procedures, this new approach may
be used by untrained personnel in everyday domestic and
clinical settings. Thus, it may represent an ecologically valid,
cost-effective, and accessible approach enabling large scale,
continuous monitoring of infant NS. This kind of studies will,
in turn, enable the assessment of the development of NS, and
its deviation in cases of pathologies.
After approval by an ethical committee, a pilot study on a
small number of newborns will be carried out to verify the
performance of the proposed technology in the field.
REFERENCES
[1] H. F. R. Prechtl, “State of the art of a new functional assessment of
the young nervous system. An early predictor of cerebral palsy,” Early
Human Develop., vol. 50, pp. 1–11, Nov. 1997.
[2] E. Thelen, “Motor development as foundation and future of developmen-
tal psychology,” Int. J. Behavioral Develop., vol. 24, no. 4, pp. 385–397,
Dec. 2000.
[3] J. M. Iverson, “ Developing language in a developing body: The
relationship between motor development and language development,”
J. Child Lang., vol. 37, pp. 229–261, Jan. 2010.
[4] C. von Hofsten, “Action in development,” Develop. Sci., vol. 10, no. 1,
pp. 54–60, Jan. 2007.
[5] D. Formica, S. K. Charles, L. Zollo, E. Guglielmelli, N. Hogan,
and H. I. Krebs, “The passive stiffness of the wrist and forearm,”
J. Neurophysiol., vol. 108, no. 4, pp. 1158–1166, 2012.
[6] D. Formica, L. Zollo, and E. Guglielmelli “Torque-dependent compli-
ance control in the joint space for robot-mediated motor therapy,” Trans.
ASME,J.Dyn.Syst.,Meas.,Control, vol. 128, no. 1, pp. 152–158, 2006.
[7] G. Pellegrino, M. Tombini, G. Assenza, M. Bravi, S. Sterzi, V. Giacobbe,
L. Zollo, E. Guglielmelli, G. Cavallo, F. Vernieri, and F. Tecchio, “Inter-
hemispheric coupling changes associate with motor improvements after
robotic stroke rehabilitation,” Restorat. Neurol. Neurosci., vol. 30, no. 6,
pp. 497–510, 2012.
[8] D. Accoto, R. Sahai, F. Damiani, D. Campolo, E. Guglielmelli, and
P. Dario, “A slip sensor for biorobotic applications using a hot
wire anemometry approach,” Sens. Actuators A, Phys., vol. 187,
pp. 201–208, Nov. 2012.
[9] M. T. Francomano, D. Accoto, E. Morganti, L. Lorenzelli, and
E. Guglielmelli, “A microfabricated flexible slip sensor,” in Proc.
IEEE RAS EMBS Int. Conf. Biomed. Robot. Biomech., Jun. 2012,
pp. 1919–1924.
[10] F. Taffoni and C. von Hofsten, “Tact glossary: Toy,” Clinica Terapeutica,
vol. 161, no. 6, pp. 573–574, 2010.
[11] D. Campolo, F. Taffoni, D. Formica, G. Schiavone, F. Keller, and
E. Guglielmelli, “Inertial-magnetic sensors for assessing spatial cog-
nition in infants,” IEEE Trans. Biomed. Eng., vol. 58, no. 5,
pp. 1499–1503, May 2011.
TAFFONI et al.: ECOLOGICAL SUCKING MONITORING OF NEWBORNS 4567
[12] D. Campolo, F. Taffoni, D. Formica, J. Iverson, L. Sparaci, F. Keller,
and E. Guglielmelli, “Embedding inertial-magnetic sensors in everyday
objects: Assessing spatial cognition in children,” J. Integr. Neurosci.,
vol. 11, no. 1, pp. 103–116, Mar. 2012.
[13] F. Taffoni, D. Formica, A. Zompanti, M. Mirolli, G. Balsassarre,
F. Keller, and E. Guglielmelli, “A mechatronic platform for behavioral
studies on infants,” in Proc. 4th IEEE RAS EMBS Int. Conf. BioRob,
Jun. 2012, pp. 1874–1878.
[14] F. Taffoni, V. Focaroli, D. Formica, J. M. Iverson, F. Keller, and
E. Gugliemelli, “Sensor-based technology in the study of motor skills
in infants at risk for ASD,” in Proc. 4th IEEE RAS EMBS Int. Conf.
BioRob, Jun. 2012, pp. 1879–1883.
[15] F. Sergi, D. Accoto, N. L. Tagliamonte, G. Carpino, and E. Guglielmelli,
“A systematic graph-based method for the kinematic synthesis of non-
anthropomorphic wearable robots for the lower limbs,” Frontiers Mech.
Eng., vol. 6, no. 1, pp. 61–70, 2011.
[16] H. Forssberg, “Neural control of human motor development,” Current
Opinion Neurobiol. vol. 9, pp. 676–682, Dec. 1999.
[17] A. Kurjak, M. Stanojevic, W. Andonotopo, A. Salihagic-Kadic,
J. M. Carrera, and G. Azumendi, “Behavioral pattern continuity from
prenatal to postnatal life a study by four-dimensional (4D) ultrasonog-
raphy,” J. Perinatal Med., vol. 32, no. 4, pp. 246–253, 2004.
[18] J. L. Miller, B. C. Sonies, and C. Macedonia, “Emergence of oropha-
ryngeal, laryngeal and swallowing activity in the developing fetal upper
aerodigestive tract: An ultrasound evaluation,” Early Human Develop.,
vol. 71, no. 1, pp. 61–87, 2003.
[19] M. Hack, M. M. Estabrook, and S. S. Robertson, “Development of
sucking rhythm in preterm infants,” Early Human Develop., vol. 11,
no. 2, pp. 133–140, 1985.
[20] P. H. Wolff, “The serial organization of sucking in the young infant,”
Pediatrics, vol. 42, no. 6, pp. 943–956, 1968.
[21] A. J. Sameroff, “The components of sucking in the human newborn,”
J. Experim. Child Psychol., vol. 6, no. 4, pp. 607–623, 1968.
[22] R. A. Waterland, R. I. Berkowitz, A. J. Stunkard, and V. A. Stallings,
“Calibrated-orifice nipples for measurement of infant nutritive sucking,”
J. Pediatr., vol. 132, no. 3, pp. 523–526, 1998.
[23] A. J. Nowak, W. L. Smith, and A. Erenberg, “Imaging evaluation of
artificial nipples during bottle feeding,” Archives Pediatr., Adolescent
Med., vol. 148, no. 1, pp. 40–42, 1994.
[24] M. E. R. Macías and G. J. S. M. Meneses, “Physiology of nutritive
sucking in newborns and infants,” Boletín Médico Hospital Infantil
México, vol. 68, no. 4, pp. 296–303, 2011.
[25] C. Lau, “Oral feeding in the preterm infant,” NeoReviews, vol. 7, no. 1,
pp. e19–e27, 2006.
[26] K. Mizuno and A. Ueda, “The maturation and coordination of sucking,
swallowing, and respiration in preterm infants,” J. Pediatr., vol. 142,
no. 1, pp. 36–40, 2003.
[27] K. Mizuno and A. Ueda, “Neonatal feeding performance as a predictor
of neurodevelopmental outcome at 18 months,” Develop. Med. Child
Neurol., vol. 47, no. 5, pp. 299–304, 2005.
[28] I. H. Gewolb, J. Bosma, E. Reynolds, and F. Vice, “Integration of suck
and swallow rhythms during feeding in preterm infants with and without
bronchopulmonary dysplasia,” Develop. Behavioral Pediatr., vol. 45,
no. 5, pp. 344–348, 2003.
[29] B. Medoff-Cooper, W. Bilker, and J. Kaplan, “Suckling behavior as
a function of gestational age: A cross-sectional study,” Infant Behavior
Develop., vol. 24, no. 1, pp. 83–94, 2001.
[30] B. Medoff-Cooper, J. Shults, and J. Kaplan, “Sucking behavior of
preterm neonates as a predictor of developmental outcomes,” J. Develop.
Behavioral Pediatr., vol. 30, no. 1, pp. 1–22, 2008.
[31] C. M. Craig, M. A. Grealy, and D. N. Lee, “Detecting motor abnor-
malities in preterm infants,” Experim. Brain Res., vol. 131, no. 3,
pp. 359–365, Apr. 2000.
[32] S. P. Costa, L. van den Engel-Hoek, and A. F. Bos, “Sucking and
swallowing in infants and diagnostic tools,” J. Perinatol., vol. 28, no. 4,
pp. 247–257, 2008.
[33] M. A. Qureshi, F. L. Vice, V. L. Taciak, J. F. Bosma, and I. H. Gewolb,
“Changes in rhythmic suckle feeding patterns in term infants in the first
month of life,” Develop. Med. Child Neurol., vol. 44, no. 1, pp. 34–39,
2002.
[34] L. Jain, E. Sivieri, S. Abbasi, and V. K. Bhutani, “Energetics and
mechanics of nutritive sucking in the preterm and term neonate,”
J. Pediatr., vol. 111, no. 6, pp. 894–898, 1987.
[35] O. P. Mathew, “Breathing patterns of preterm infants during bottle
feeding: Role of milk flow,” J. Pediatr., vol. 119, no. 6, pp. 960–965,
Dec. 1991.
[36] M. Taki, K. Mizuno, M. Murase, Y. Nishida, K. Itabashi, and Y. Mukai,
“Maturational changes in the feeding behaviour of infants—A com-
parison between breast-feeding and bottle-feeding,” Acta Paediatrica,
vol. 99, no. 1, pp. 61–67, 2010.
[37] S. Fadavi, I. C. Punwani, L. Jain, and D. Vidyasagar, “Mechanics and
energetics of nutritive sucking: A functional comparison of commer-
cially available nipples,” J. Pediatr., vol. 130, no. 5, pp. 740–745, 1997.
[38] W. C. Lang, N. R. M. Buist, A. Geary, S. Buckley, E. Adams,
A. C. Jones, and S. Gorsek, “Quantification of intraoral pressures during
nutritive sucking: Methods with normal infants,” Dysphagia, vol. 26,
no. 3, pp. 277–286, 2011.
[39] S. Cecchini, E. Schena, and S. Silvestri, “An open-loop controlled active
lung simulator for preterm infants,” Med. Eng. Phys., vol. 33, no. 1,
pp. 47–55, 2011.
[40] C. Lau and R. J. Schanler, “Oral feeding in premature infants: Advantage
of a self-paced milk flow,” Acta Paediatrica, vol. 89, no. 4, pp. 453–459,
2000.
[41] E. Tamilia, F. Taffoni, E. Schena, D. Formica, L. Ricci, and
E. Guglielmelli, “A novel ecological method for the estimation of
nutritive sucking effiviency in newborns: Measurement principle and
experimental assessment,” in Proc. Int. Conf. IEEE EMBC, Jul. 2013,
pp. 6720–6723.
Fabrizio Taffoni (M’11) received the Ph.D. degree
in biomedical engineering from Università Cam-
pus Bio-Medico di Roma, Rome, Italy, in 2009.
From 2009 to 2012, he was a Post-Doctoral Fellow
with the Laboratory of Biomedical Robotics and
Biomicrosystems, Università Campus Bio-Medico
di Roma, where he has served as a Tenure Track
Assistant Professor of bioengineering since 2012.
His current research interests include intersection
between developmental neuroscience, bioengineer-
ing, and mechatronics, with a focus on the design
and development of new platforms, tools, and methods for ecological assess-
ment of motor development.
Eleonora Tamilia (M’13) received the M.S. degree
in medical engineering from the University of Rome
“Tor Vergata,” Rome, Italy, in 2010. She is currently
pursuing the Ph.D. degree in biomedical engineer-
ing with Università Campus Bio-Medico di Roma,
Rome, where she joined the Laboratory of Biomed-
ical Robotics and Biomicrosystems in January 2012.
Her current research interests include devices and
methods for infants’ neurodevelopment assessment,
with a focus on newborns’ sucking behavior, and
analysis of experimental data on children’s behavior
during neurodevelopment.
Maria Rosaria Palminteri received the master’s
degree in biomedical engineering from Università
Campus Biomedico di Roma, Rome, Italy, in 2012.
She was with the Laboratory of Biomedical Robotics
and Biomicrosystems, focusing on the design and
development of new methods and tools for non-
obtrusive assessment of suction in infants. She is
currently a Research and Development Engineer
with CRAD Imaging AB, Uppsala, Sweden. Her
current research interests include the medical devices
industry and engineering design.
4568 IEEE SENSORS JOURNAL, VOL. 13, NO. 11, NOVEMBER 2013
Emiliano Schena (M’10) received the Ph.D. degree
in biomedical engineering from Università Campus
Bio-Medico di Roma, Rome, Italy, in 2009. He is
currently a Researcher of mechanical and thermal
measurements with Università Campus Bio-Medico
di Roma. His current research interests include sen-
sors and transducers for mechanical and thermal
measurements in the biomedical and clinical field,
fiber optic-based measurement systems, and respira-
tory rehabilitation.
Domenico Formica (M’10) received the B.S. and
M.S. degrees in biomedical engineering and the
Ph.D. degree in bioengineering from Università
Campus Bio-Medico di Roma, Rome, Italy, in 2002,
2004, and 2008, respectively. From 2008 to 2010,
he was a Post-Doctoral Fellow with the Laboratory
of Biomedical Robotics and Biomicrosystems, Uni-
versità Campus Bio-Medico di Roma, where he is
currently a Tenure-Track Assistant Professor of bio-
engineering. His current research interests include
mechatronic technologies for the study of human
motor control, with particular attention to neurodevelopment, quantitative
algorithms for clinical assessment of patients with neuromuscular disorders,
and novel robotic devices for rehabilitation motor therapy after neurological
injury, with special attention to the issue of interaction control.
Jonathan Delafield-Butt received the Ph.D. degree
in developmental neurobiology from the University
of Edinburgh Medical School, Edinburgh, Scotland,
in 2003. He is currently a Lecturer with the Uni-
versity of Strathclyde, Glasgow, U.K. His current
research interests include early infant psychomo-
tor development with applications in improving
understanding and treatment of developmental psy-
chopathology, socioemotional care, and learning.
Flavio Keller received the Postgraduate degree in
neuropharmacology from the University of Zurich,
Zurich, Switzerland, in 1982. He is currently a Full
Professor of physiology with Università Campus
Bio-Medico di Roma, Roma, Italy. He is the Head
of the Laboratory of Developmental Neuroscience
and Neural Plasticity. His current research interests
include developmental neuroscience and neurobiol-
ogy (specifically the neurobiology of sensorimotor
systems), medicine, and biochemistry. He is the
author or co-author of more than 80 works in peer-
reviewed international journals, conference proceedings, and books.
Sergio Silvestri (M’01) received the Ph.D. degree in
mechanical and thermal measurements from Cagliari
University, Cagliari, Italy, in 2001. He is currently
an Associate Professor of measurements and bio-
medical instrumentation with the Università Cam-
pus Bio-Medico di Roma, Rome, Italy. His current
research interests include sensors for biomedical
devices and instrumentation, and minimally invasive
and non-invasive medical measurement systems.
Eugenio Guglielmelli (SM’11) received the Degree
in electronics engineering and the Ph.D. degree in
biomedical robotics from the University of Pisa,
Pisa, Italy, in 1991 and 1995, respectively. He is cur-
rently a Full Professor of bioengineering with Uni-
versità Campus Bio-Medico di Roma, Rome, Italy,
where he serves as the Head of the Laboratory of
Biomedical Robotics and Biomicrosystems, which
he founded in 2004. His current research inter-
ests include human-centred robotics, biomechatronic
design and biomorphic control of robotic systems,
and their application to robot-mediated motor therapy, assistive robotics, and
neurorobotics. He is the author or co-author of more than 170 papers in
peer-reviewed international journals, conference proceedings, and books. He
is a co-inventor of three patents and co-founder of four spin-off companies.
He served as an Associate Editor of the IE EE ROBOTICS &AUTOMATION
MAGAZINE and he serves as an Associate Editor of the IE EE TRANSAC-
TIONS ON ROBOTICS. He serves as an Editor-in-Chief of the Springer Series
on Biosystems and Biorobotics.
