Content uploaded by Yuri D Feldman
Author content
All content in this area was uploaded by Yuri D Feldman
Content may be subject to copyright.
Human Skin as Arrays of Helical Antennas in the Millimeter and Submillimeter Wave Range
Yuri Feldman,
1,
*Alexander Puzenko,
1
Paul Ben Ishai,
1
Andreas Caduff,
1
and Aharon J. Agranat
1,2
1
Department of Applied Physics, The Hebrew University of Jerusalem,
Givat Ram, 91904, Jerusalem, Israel
2
The Interdisciplinary Center for Neural Computation, The Hebrew University of Jerusalem,
Givat Ram, 91904, Jerusalem, Israel
(Received 22 August 2007; published 27 March 2008)
Recent studies of the minute morphology of the skin by optical coherence tomography showed that the
sweat ducts in human skin are helically shaped tubes, filled with a conductive aqueous solution. A
computer simulation study of these structures in millimeter and submillimeter wave bands show that the
human skin functions as an array of low-Qhelical antennas. Experimental evidence is presented that the
spectral response in the sub-Terahertz region is governed by the level of activity of the perspiration
system. It is also correlated to physiological stress as manifested by the pulse rate and the systolic blood
pressure.
DOI: 10.1103/PhysRevLett.100.128102 PACS numbers: 87.50.S, 87.90.+y
Experimental evidence indicating that the electromag-
netic properties of the human skin in the sub-Terahertz
frequencies are governed by its morphology is henceforth
presented.
The human skin is the largest organ of the body, de-
signed as the primary interface, utilizing numerous of
functions and interactions between us and our environ-
ment. The complexity of the multilayered skin morphology
provides an extremely broad range of features of sensors
that utilize a number of physical phenomena. One of these
skin features is the perspiration system that traditionally is
mainly considered for body thermoregulation [1]. Its main
components are sweat glands embedded into the dermis
and connecting through the epidermis with the pores on the
surface of the stratum corneum by ducts, filled with a
conductive aqueous solution. The general illustration of
sweat glands presents a convoluted arrangement for the
sweat gland and a more or less straight tube for the duct
[1,2]. In recent investigations of the subcutaneous mor-
phology of the human skin by optical coherent tomography
[3,4], it was found that the sweat duct is in fact a remark-
ably arranged helical conductive tube (Fig. 1). This, to-
gether with the fact that the dielectric permittivity of the
dermis is higher than that of the epidermis, brings forward
the supposition that as electromagnetic entities, the sweat
ducts could be regarded as low-Qhelical antennas.
Inherent to this supposition is the requirement that the
duct possesses an electrical conductance mechanism that
is effective at the extremely high frequency (EHF) range.
Even though the ducts are filled with conducting electro-
lytes, the ions mobility rates associated with sweat are slow
compared to the characteristic frequencies under consid-
eration. A mechanism that qualifies for such a requirement
is fast proton hopping through distributed H-bond net-
works along the duct surface. It is well established that
these networks exist in biological structures [5] and it was
found that the characteristic time for such proton transport
is about 1013 sec [6].
When the potential drop caused by the difference in pH
values between the skin surface and the dermis is taken into
consideration [2], such hopping can account for the ac
conductivity that is necessary for the sweat ducts to yield
an electromagnetic response in the EHF range. Moreover,
it is known that the human skin contains approximately 2 to
5106eccrine sweat glands distributed over most of the
body, with higher density in several areas such as on the
palms of the hand, the forehead, and on the soles of the feet
[7,8]. As each gland is connected to the skin surface by a
helical sweat duct, the skin organ in its entirety can be
regarded as an array of helical antennas that operate in the
EHF range. It has been ascertained that the level of sweat-
ing has a dominant effect on the conductance parameters of
the various components of the skin tissue. As pointed out
above, these parameters strongly affect the spectral re-
sponse of the skin organ. Hence, it is predicted that the
physiological and psychological parameters that are known
to be expressed in the activity of the perspiration system [9]
FIG. 1. 3D optical coherence tomography image (reproduced
with permission from ISIS GmbH) of a single human eccrine
sweat gland embedded in the human skin and a schematic
presentation of the duct as a helical antenna [20] embedded in
the skin, where the dermis-epidermis interface acts as a dielectric
reflector. The respective permittivities of the skin layers are
marked. They were estimated for the specific frequency range
based on the water content of the layers [10].
PRL 100, 128102 (2008) PHYSICAL REVIEW LETTERS week ending
28 MARCH 2008
0031-9007=08=100(12)=128102(4) 128102-1 ©2008 The American Physical Society
will also be manifested in the spectral response of the skin
in the EHF frequencies.
To test the validity of this prediction we conducted a
series of in vivo measurements of the skin reflection co-
effcient of the hand palm in several subjects. The first set of
measurements was done using a vector network analyzer
(VNA) in the spectral range of 75 GHz to 110 GHz (W
band). In order to avoid parasitic reflections and diffraction
effects the first set of measurements was done in a system
that was configured for near field measurements. The
initial real system measurement was done when the subject
was fully rested. The subsequent measurement was done
immediately after a period of 20 min of intense jogging,
and was followed by a series of measurements every 1 min
as the subject relaxed. A typical series of measurements for
one subject is presented in Fig. 2(a). As can be seen there is
a pronounced difference between the reflectance spectra
that were measured in the calm state and immediately after
a period of physical activity. In the subsequent measure-
ments, as the subject relaxed back to the calm state, the
spectral response of the coefficient also relaxed back to-
wards its initial curve. These results were compared to a
computational study of the propagation of an electromag-
netic beam in an idealized section of the skin containing
eight helical sweat ducts. The sweat ducts were modeled as
conducting coils of 2 to 4 turns with diameters 60 to
80 mand heights 300 to 350 m(see Figs. 3and 4).
The skin layer was modeled as three strata representing the
stratum corneum, the epidermis, and the dermis. The water
concentration was set to be 10%–15% in the stratum
corneum, 45%–55% in the epidermis, and 70%– 80% in
the dermis [10]. The stability of the hydrogen network
inside the coil and hence the resultant level of proton
conductivity is expected to exhibit a direct dependence
on the sweat rate. Hence, the conductivity of the coil can
be used as the parameter that quantifies the level of relaxa-
tion following intense physical activity [11,12], i.e., it is
expected that the conductivity of the coil will follow the
same time dependence as the decaying sweat flow in the
ducts. The computed spectra of the reflectance coefficient
for different levels of duct conductivity are shown in
Fig. 2(b). It can be seen that the spectral response of the
reflectance coefficient is very similar to the experimental
results.
The second set of measurements was done in a system
that was configured for distance measurements. The palm
was held steady by a stand that was placed 22 cm from the
horn antenna at the input of the VNA, and a dielectric lens
was used to collate the beam. Sets of identical measure-
ments were taken of an ensemble of 13 subjects differing in
gender, age and ethnic origin. Each set included a mea-
surement of the skin reflectance, and concurrent recordings
of the pulse rate, the systolic blood pressure, and the skin
temperature.
The subjects performed 20 min of jogging after which a
sequence of 30 sets of measurements were taken at 1 min
intervals.
A typical sequence of sets of measurements is presented
in Fig. 5(a). The skin reflectance is presented in terms of its
FIG. 2 (color). (a) Measurements of the modulus of the reflec-
tion coefficient Rof the human palm in the frequency region 75
to 110 GHz. The subject was measured using VNA HP 8510C in
the near field in a calm state and then during 30 min relaxation
after intense physical activity. The arrow on the graph indicates
the direction of the time line and shows how the signal returns to
the calm state with relaxation; (b) The simulation of the reflec-
tion coefficient from 8 coils embedded in an idealized skin.
FIG. 3. The model used for 3D Electromagnetic simulations.
The disk represents a portion of skin consisting of 3 separate
layers and an array of 8 sweat ducts being subjected to a signal
from the wave guide.
PRL 100, 128102 (2008) PHYSICAL REVIEW LETTERS week ending
28 MARCH 2008
128102-2
frequency averaged relative signal intensity given by
hWreli 1
f2f1Zf2
f1
jUsubjectj2
jUreferencej2df; (1)
where Usubjectfis the reflected signal from the subject
after physical activity, Ureferencefis the reflected signal
measured from the subject while sitting calmly before
physical activity, f175 GHz, and f2110 GHz. It can
be seen that after the physical activity an exponential-like
relaxation is observed, which is correlated to physiological
stress as evidenced by parallel relaxations in pulse rate and
systolic blood pressure. The results are summarized in
Fig. 5(b) in which the normalized ensemble average of
Wrel, denoted as hWrel i, for the 30 measurement points are
presented vs the respective ensemble average of the sys-
tolic blood pressure. A strong correlation, defined by the
coefficient ris clearly manifested with r0:984
rPN
i1xi
xyi
y
PN
i1xi
x2PN
i1yi
y2
q:(2)
To rule out the possibility that the observed phenomena
could be due to the water content in the skin and underlying
tissues, an additional set of measurements using a pressure
cuff were performed. This allowed control of the blood
flow during the measurement without activating the sweat
gland system. As the cuff pressure is increased (0 –100 mm
Hg), it reduces the capillary blood flow [13] resulting in an
increase of the total amount of blood in the skin and
underlying tissue [14].
Measurements of the normalized average reflectance
show no noticeable dependence on the capillary blood flow
or change in volume fraction in this tissue compartment.
In order to test the effect of active or inactivate sweat
glands on the reflection coefficient a creme containing a
FIG. 4. The idealized sweat duct used in the simulation soft-
ware ’’CST microwave studio’’ with the relevant dimensions.
The sweat duct coil was modeled as a helical pipe filled with
electrolyte. It is permanently full of sweat and so there exists a
hydrogen bond network along the surface. Fast Proton hopping
in H-bond network has been measured at 1013 [6]. The differ-
ence in pHbetween the skin surface and the dermis results in a
concentration gradient H3106mole=land subse-
quent potential drop. This is the possible cause of fast currents
in the coil. Proton conductivity in bulk water in biological
structures has been measured at 100–1000 S=m[21].
Therefore in the simulation duct conductivity was set accord-
ingly high, 500–20 000 S=m[see Fig. 2(b)].
FIG. 5 (color). (a) The frequency averaged relative signal
intensity hWrelirecorded from reflection coefficient measure-
ments in the frequency band 75–110 GHz of the palm of a
subject at rest following 20 min of intense physical activity. To
avoid reflection from surface water the hand was kept dry; (b) the
correlation graph between systolic blood pressure and hWrel i,
obtained from the average of 13 subjects measured as they
relaxed for a period of 30 min after intensive physical activity.
The intensities and blood pressures were normalized over their
amplitudes to allow averaging and the correlation coefficient was
calculated from linear regression. The value r0:984, close to
unity, demonstrates a strong correlation between them.
Essentially they exhibit similar temporal behavior. The correla-
tion of hWreliwith the pulse rate is r0:85.
PRL 100, 128102 (2008) PHYSICAL REVIEW LETTERS week ending
28 MARCH 2008
128102-3
snake venomlike synthetic tripeptide acting as an antago-
nist of the postsynaptic muscular nicotinic acetylcholine
membrane’s receptor (mnAChR), was applied to the test
area [15]. The measurements of the reflection coefficient
were repeated again after exercise. The same subjects were
then treated with a placebo creme, based on the same
matrix but not containing the synthetic tripeptide [16].
This was done in order to account for any hydrating ef-
fects of the creme itself. The results demonstrated a sig-
nificantly lowered signal intensity when the active compo-
nent was used, indicating the importance of neurally
active/regulated sweat glands in the received signal. The
results are illustrated in Fig. 6.
In summary it is claimed that individual sweat ducts are
low-Qhelical antennas and that their presence in the skin
means that the skin can be regarded as a 2D antenna array
in the sub-terahertz region. The spectral response is sensi-
tive to the activity of the sweat system. These claims were
substantiated experimentally where it was shown that the
spectral response of the EM reflectance of the skin is
indeed correlated with the activity level of the perspiration
system and follows the same temporal behavior as other
physiological parameters, such as the pulse rate and the
systolic blood pressure. This phenomenon can be used as
the basis for a generic remote sensing technique for pro-
viding a spatial map of the sweat gland activity of the
examined subjects. As the mental state and sweat gland
activity are correlated [17–19] it has the potential to be-
come a method for providing by remote sensing informa-
tion regarding some physiological parameters and the
mental state of the patients.
We thank Professor B. Kapilevich at Ariel College for
helpful discussions, Professor D. Davydov, Dr. M.
Golosovski and Mr. F. Sakran of the Physics Department
of HUJI for their assistance, Mr. E. Polygalov for stimu-
lating discussions, Dr. A. Greenbaum, Miss O. Heller,
Mr. I. Davidovich, and other members of the Dielectric
Spectroscopy Laboratory of the Department of Applied
Physics HUJI for their comprehensive help in the VNA
measurements. This work was supported by grants from
the Israel Science Foundation Grant No. 1128/05 and the
Yeshaya Horowitz Association.
*Author to whom correspondence should be addressed.
yurif@vms.huji.ac.il
[1] H. R. Jacubovic and A. B. Ackerman, Dermatology (W. B.
Saunders, Philadelphia, 1992), 3rd ed., pp. 66 –77.
[2] Textbook of Medical Physiology (Sanders, Philadelphia,
1991), 8th ed.
[3] A. Knu
¨ttel and M. Bo
¨hlau-Godau, J Biomed. Opt. 5,83
(2000).
[4] A. Knu
¨ttel et al., J Biomed. Opt. 9, 265 (2004).
[5] R. R. Sadeghi and H. P. Cheng, J. Chem. Phys. 111, 2086
(1999).
[6] V. V. Krasnoholovets et al., Adv. Chem. Phys. 125, 351
(2003).
[7] T. Nishiyama et al., Auton. Neurosci. Basic 88, 117
(2001).
[8] R. Salvesen, J. Neurol. Sci. 183, 39 (2001).
[9] T. Ogawa and J. Sugenova, J. Physiol. Soc. Jpn. 43, 275
(1993).
[10] P. J. Caspers et al., J. Raman Spectrosc. 31, 813 (2000).
[11] R. Perini et al., Eur J. Appl. Physiol. 58, 879 (1989).
[12] Y. Inoue, Eur. J. Appl. Physiol. 79, 17 (1998).
[13] A. Murray and D. Marjanovic, Med. Biol. Eng. Comput.
35, 425 (1997).
[14] L. Douven and G. W. Lucassen, Proc. SPIE Int. Soc. Opt.
Eng. 3914, 312 (2000).
[15] J. J. Mcardle et al., J. Pharmacol. Exp. Therap. 289, 543
(1999).
[16] H. Ziegler and M. Heidl, Cosmet. Toiletries 122,59
(2007).
[17] D. M. Dipasquale et al., J. Physiol. Soc. Jpn. 53, 427
(2003).
[18] A. K. M. Shamsuddin and T. Togawa, Physiol. Meas. 21,
535 (2000).
[19] S. C. Landis and D. Keefe, Dev. Biol. 98, 349 (1983).
[20] A. Balanis, Antenna Theory: Analysis and Design (Wiley,
New Jersey, 2005).
[21] S. Cukierman, Biophys. J. 78, 1825 (2000).
FIG. 6. The effect of temporally deactivated sweat glands on
the relative signal intensity is illustrated by the lowered ampli-
tude, using the synthetic tripeptide, applied to the skin surface
120 min before measurement. The amplitudes are averaged over
8 subjects and shown relative to the amplitude recorded when
using a placebo creme.
PRL 100, 128102 (2008) PHYSICAL REVIEW LETTERS week ending
28 MARCH 2008
128102-4
