COMPARISON BETWEEN LASER DOPPLER

FLOWMETRY SIGNALS RECORDED IN GLABROUS

AND NON GLABROUS SKIN

Time and Frequency Analyses

Edite Figueiras

Instrumentation Center (GEI-CI), Physics Department, Faculty of Sciences and Technology of Coimbra University

Rua Larga, P-3004-516 Coimbra, Portugal

Pierre Abraham

Laboratoire de Physiologie et d’Explorations Vasculaires, UMR CNRS 6214-INSERM 771

Centre Hospitalier Universitaire d’Angers, 49033 Angers Cedex 01, France

Luis F. Requicha Ferreira

Instrumentaiton Center (GEI-CI), Physics Department, Faculty of Sciences and Technology of Coimbra University

Rua Larga, P-3004-516 Coimbra, Portugal

Anne Humeau

Groupe Esaip, 18 rue du 8 mai 1945, BP 80022, 49180 Saint Barthélémy d’Anjou cedex, France

Laboratoire d’Ingénierie des Systèmes Automatisés (LISA), Université d’Angers

62 Avenue Notre Dame du Lac, 49000 Angers, France

Keywords: Laser Doppler flowmetry, Skin microcirculation, Glabrous skin, Non glabrous skin, Power-law, Fractal.

Abstract: Skin microvascular properties vary with anatomical zones. Thus, glabrous skin found in fingers, toes, nail

beds, hand palms and feet soles has a high density of arteriovenous anastomoses (AVAs). In contrast, skin

found in sites such as ventral face of the forearms do not possess AVAs and therefore microvascular blood

flow in this non glabrous skin is different. We herein propose to analyse laser Doppler flowmetry (LDF)

signals that reflect skin microvascular perfusion, in two different sites of healthy subjects: hand (glabrous

skin) and ventral face of the forearm (non glabrous skin). The signal analysis is performed both in the time

and in the frequency domains. Our results show that the mean amplitude of LDF signals recorded in the

hand is generally higher than in the forearm. Moreover, the signal fluctuations observed in the hand are

much higher than the ones observed in the forearm. Our work also shows that the power spectrum of LDF

signals recorded in hand and forearm can be different. They both may possess characteristics of fractal

processes but these characteristics may be different for the two anatomical sites.

1 INTRODUCTION

The human skin anatomy and function vary with age

and region of the body. Human skin consists of three

main layers: the epidermis, the dermis, and the

hypodermis. The dermis has a microvascular

network, i.e. it has blood flow passing through

vessels smaller than 100 µm (Morales, 2005),

organized in two horizontal plexuses: the upper

horizontal plexus and the lower horizontal plexus.

Some parts of the skin also possess arteriovenous

anastomoses (AVAs) or shunts that allow blood flow

to bypass superficial skin layers, thus providing

efficient thermal regulation (Berardesca et al.,

133

Figueiras E., Abraham P., Requicha Ferreira L. and Humeau A. (2010).

COMPARISON BETWEEN LASER DOPPLER FLOWMETRY SIGNALS RECORDED IN GLABROUS AND NON GLABROUS SKIN - Time and

Frequency Analyses.

In Proceedings of the Third International Conference on Bio-inspired Systems and Signal Processing, pages 133-138

DOI: 10.5220/0002588401330138

 SciTePress

2002). Therefore, different types of skin are found.

Thus, glabrous skin is found in regions such as hand

palms (where there are AVAs), whereas non

glabrous skin is found in zones such as

forearms (where there are no AVAs). Glabrous skin

is mediated by a vasoconstrictor system, whereas

non glabrous skin is mediated by both adrenergic

vasoconstrictor nerves and an active vasodilator

system.

Disorders of the blood microcirculation system

are known to play a significant role in the

development of various diseases, such as diabetes,

peripheral vascular disease or Raynaud’s

phenomenon. Because the skin is so accessible, there

are many new ways of studying it, based mainly on

the quantification of its optical and thermal

properties which are modified by the amount of

blood perfusion (Berardesca et al., 2002). These

techniques are being improved constantly. In the last

few years, attention has been drawn to the laser

Doppler flowmetry (LDF).

LDF is a non invasive technique enabling the

monitoring of microvascular blood flow. With this

technique it is possible to monitor blood flow at a

single point (laser Doppler perfusion monitoring -

LDPM) or map tissue blood perfusion (laser Doppler

perfusion imaging - LDPI) (Nilsson et al., 2003).

LDF can be used in dermatology to assess the

degree of skin irritability in patch test procedures, in

pharmacology to study the microvascular effects of

vasoactive substances and drugs, in plastic surgery...

The technique also allows the study of the changes

in microvascular blood flux in diabetic patients, in

people with peripheral vascular diseases or

Raynaud’s phenomenon.

Differences in vascular anatomy and

physiological control, and differences in scattering

and absorbency properties, have mainly been studied

with LDPM signals in different tissues, such as

muscle, liver, and skin in general. However, no deep

studies have been conducted in order to know how

glabrous and non glabrous skin properties influence

LDF recordings. Moreover, to the best of our

knowledge, no spectral domain comparison of LDF

signals recorded in hand palm (glabrous skin) and

forearm (non glabrous skin) has been performed.

However, the physiology and skin thickness of these

two regions of interest are very different. How do

these differences impact LDF recordings? In order to

answer this question, we herein propose to compare

LDF signals recorded simultaneously in glabrous

and non glabrous zones of healthy subjects. This

comparison is performed through both temporal and

spectral analyses.

2 LASER DOPPLER

FLOWMETRY

As mentioned previously, in the last years, LDF has

drawn much attention for the monitoring of skin

perfusion. In LDF technique, a coherent light beam

is directed toward the tissue under study. There, it is

scattered by moving objects and by static tissue

structures. When light is scattered by a moving

particle, like a red blood cell, it is frequency shifted.

This shift depends on the velocity of the particle, the

direction of the incoming light and the direction of

the scattered light. On the contrary, light scattered

by static structures remains unshifted in

frequency (see for example Fredriksson et al., 2007).

Thus, when a photon encounters a particle moving

with a velocity

(m/s), and if

(rad/m)

describes

the propagation vector of the incoming photon, the

propagation vector of the photon after being

scattered,

, comes out as represented in Figure 1.

The angular Doppler frequency shift w (rad/s) is:

cos

sin

)(

⎟

⎠

⎞

⎜

⎝

⎛

−=−⋅−=⋅−= vkkvvqw

(1)

where λ

represents the wavelength (m) of the photon

in the surrounding medium, α is the scattering angle

between

and

, and θ is the angle between the

projection of

in the plane of scattering

and

(

)

− vector. The difference between

and

is often denoted by the scattering vector

If the reflected mixed light (frequency shifted

and unshifted) by the skin is detected by a photode-

tector, optical mixing of light shifted and unshifted

Figure 1: Single scattering event between a photon and a

moving scatterer, in this case a red blood cell.

and

denote the incoming and scattered wave vectors, and α is

the angle between the two.

is the velocity vector of the

red blood cell.

is the difference between

and

. θ

is the angle between

and

BIOSIGNALS 2010 - International Conference on Bio-inspired Systems and Signal Processing

134

in frequency will result in a stochastic photocurrent.

The photocurrent consists of a static part and a

fluctuating part. The total signal can be described

with the autocorrelation function (ACF), which is

directly related to the power spectral density of the

signal. The autocorrelation obtained can be divided

into different terms of the origin of the current:

stationary (currents produced by the unshifted light),

heterodyne-mixing (produced by mixing of the

unshifted light and the shifted light) and homodyne-

mixing (produced by mixing of the shifted light by

red blood cells (RBCs) with different velocities).

Usually the homodyne part is disregarded, because

the measurements are made in low to moderate

blood volumes, where the heterodyne part dominates

over the homodyne part. The ACF of the heterodyne

part can be expressed as (see for example

Fredriksson

et al., 2007):

(

)

iqvtiqvt

eeICMBCACF

−

+∗=

(2)

where

I is the average of the current produced by the

unshifted light – DC current, CMBC is the

concentration of moving blood cells,

q is the

scattering vector and

v is the velocity. According to

the Wiener-Khintchine theorem, the Fourier

transform of the ACF is equal to the power spectral

density

P(w) of the input. Therefore, for the

heterodyne part we have (see for example

Fredriksson

et al., 2007):

()

dteeeICMBCwP

jwkiqvtiqvt −

∞

∞−

−

⋅+∗=

∫

)( (3)

where w is the angular frequency, k is the wave

number, v is the velocity of the RBC and q is the

scattering vector. The photocurrent power spectrum

is related to the properties of the blood cells in the

illuminated volume. By further derivation of this

expression it can be shown that the CMBC and the

perfusion (PERF) can be estimated from the power

spectrum. The CMBC is proportional to the integral

of the Doppler power spectrum density (see for

example Fredriksson et al., 2007):

∫

∞

)( dwwPCMBC

(4)

and the perfusion, in arbitrary units (a.u.) is

proportional to the integral of the frequency-

weighted Doppler power spectrum (see for example

Fredriksson et al., 2007):

∫

∞

∝∗=

)( dwwwPvCMBCPERF

(5)

where

is the mean speed of the blood cells (see

for example Fredriksson et al., 2007).

Currently, LDF does not give any absolute

measure of blood perfusion. In the clinical setting

this is a limiting factor and the reason why LDF

instruments are not routinely used in health care.

However, LDF has found its use in research.

The dynamics of the microcirculatory flow,

measured by LDF, consists of rhythmic oscillations.

The latter can be analysed using spectral techniques.

Thus, the spectral analysis of LDF signals revealed

six peaks within the range frequency from 0.005 Hz

to 2.0 Hz (see among others Stefanovska et al.,

1999). The peak in the interval from 0.6 Hz to

2.0 Hz is due to heart beats; the one between

0.145 Hz and 0.6 Hz is due to respiratory activity;

the one between 0.052 Hz and 0.145 Hz is the

intrinsic myogenic activity. The one from 0.021 Hz

to 0.052 Hz is due to the neurogenic activity caused

by the sympathetic system, whereas the one from

0.0095 Hz to 0.021 Hz is due to NO-dependent

endothelial activities. Finally, the one between

0.005 Hz and 0.0095 Hz is due to non NO-

dependent endothelial activities.

3 PHYSIOLOGICAL

DIFFERENCES BETWEEN

ARM (NON GLA-BROUS) AND

HAND (GLABROUS) SKIN

There are large differences between the

microcirculatory system in hands (glabrous skin)

and in arms (non glabrous skin). The main

difference is the high density of AVAs or shunts in

the glabrous skin (like fingers and toes, the nail

beds, the palm of the hands, the sole of the feet, and

the earlobe) that is not present in skin

elsewhere (Roustit et al., 2008). However, portion of

cardiac output passing through skin AVAs in

humans is not well known. Regarding the reflex

control of the skin blood flow, there are differences

when glabrous and non glabrous skin are compared:

non glabrous skin is mediated by both adrenergic

vasoconstrictor nerves and an active vasodilator

system, whereas glabrous skin is mediated by a

vasoconstrictor system only - the classic adrenergic

nervous system (Wilson et al., 2005).

COMPARISON BETWEEN LASER DOPPLER FLOWMETRY SIGNALS RECORDED IN GLABROUS AND NON

GLABROUS SKIN - Time and Frequency Analyses

135

Very few studies have compared LDF signals

from fingers and arms. A characteristic pattern of

large, spontaneous fluctuations in blood flow has

been described in human glabrous skin by several

authors (see for example Thoresen and Walloe,

1980). Some authors assumed that the fluctuations

are caused by synchronous opening and closing of

skin AVAs (Lossius et al., 1992). Skin AVAs are

densely innervated with sympathetic vasoconstrictor

fibres. There is also a connection between the

fluctuations in glabrous skin blood flow and the

spontaneous heart rate and blood pressure

variability (Lossius et al., 1992).

Moreover, measurements were made both on the

fingertip and in dorsal forearm skin by Freccero et

al. (2003) and it was concluded that local heating

increases superficial blood flow in fingertip and

forearm skin by different adjustment of blood cell

concentration and velocity (differences are of a

rather minor character). Roustit et al. (2008) studied

the lidocaine/prilocaine effect, a pharmacological

tool to inhibit the axon reflex, on finger pads and

forearms when they are submitted to a local heating.

They found that there is a smaller effect of

lidocaine/prilocaine cream on the finger pad. This is

due to a decreased anesthetic effect of topical

lidocaine/prilocaine on the finger pads. In another

study, conducted by Roustit et al. (2009), the sodium

nitroprusside (SNP) iontophoresis test, used to

assess the non endothelium-dependent

microvascular function of the finger pad, was

compared with SNP iontophoresis test on the

forearm, because most data available on SNP

iontophoresis concerns the skin of the forearm. In

the forearm there was an increase in cutaneous

vascular conductance but, on the finger pad, such

hyperemia was not consistent. They concluded that

standard protocols used for SNP iontophoresis

cannot be used on the finger pad as tools to assess

non-endothelium-dependent skin microvascular

dilation. Also, the thicker epidermis of the finger

pulp may present a barrier to the diffusion.

Furthermore, the effect of nerve blockade on

forearm and finger skin blood flow during body

heating and cooling was studied by Saumet et

al. (1992). They concluded that the active

vasodilator system plays an important role, as far as

the timing and the amplitude of the cutaneous

vasodilator response to whole body heating in the

forearm, but not in the finger. The vasoconstrictor

response to cooling occurred only in the finger.

Moreover, there were found differences in

vasodilator response in the two types of skin (Tucker

et al., 1998). The authors attributed these differences

to the higher baseline flow in the finger circulation.

Wilson et al. (2005) tested the hypothesis that,

independent of neural control, glabrous and non

glabrous cutaneous vasculature is capable of

autoregulating blood flow. In addition to neural

control, a number of local factors are capable of

modulating skin blood, for example, local alterations

in temperature and venous congestion or increased

transmural pressure. They found that glabrous skin

of the hand palm has the capability to autoregulate

blood flow in response to dynamic changes in blood

pressure. However, they noted less intrinsic

autoregulatory capabilities in non glabrous skin of

the forearm. Glabrous skin is capable of both static

and dynamic autoregulation while non glabrous skin

retains static with no dynamic autoregulatory

capabilities.

4 COMPARISON BETWEEN LDF

SIGNALS RECORDED IN

GLABROUS AND NON

GLABROUS SKIN

4.1 Measurement Procedure

In order to compare LDF signals from glabrous and

non glabrous skin, thirteen healthy subjects (between

21 and 44 years old) were studied. All were

informed of the measurement procedure and gave

their written informed consent. After at least 10 min

of acclimatization in the supine position, the

acquisitions started in a room at ambient

temperature. The flowmeter used was a Periflux

5000 (Perimed, Sweden) for which the time constant

was chosen equal to 0.2 s and the wavelength was

780 nm. Two signals were recorded simultaneously

in a.u.: one probe of the flowmeter was positioned in

the ventral face of the right forearm of the

subject (non glabrous skin), while the other probe

was positioned in the right hand palm (glabrous

skin). The two signals were recorded for at least

5 min with a sampling frequency of 20 Hz. In what

follows, 6000 samples of signals (5 min) are

processed. Two signals recorded simultaneously in

the forearm and in the hand are shown in Figure 2.

4.2 Signal Processing Analysis

In what follows, temporal and spectral analyses of

LDF signals recorded simultaneously in glabrous

and non glabrous skins are performed. For the

spectral study, the power spectrum is computed and

possible power-law properties are analyzed.

BIOSIGNALS 2010 - International Conference on Bio-inspired Systems and Signal Processing

136

0 50 100 150 200 250 300

100

150

200

Time (s)

Solid line: forearm; dotted line: hand

Figure 2: LDF signals recorded simultaneously in a

healthy subject. The red solid line (lower curve)

corresponds to the data recorded in the ventral face of the

forearm; the black line (upper curve) corresponds to the

data recorded in the hand palm.

Power-law relationship is observed when the

general shape of the power spectrum is a power-law

decreasing curve without eminent peak. In that case,

the power versus frequency relationship is:

()

−

ffPS ~

(6)

where PS is the power spectrum and f is the

frequency. In a log-log plot, Eq. 6 gives a straight

line with slope β. For such signals, data may be

regarded as a fractal: the corresponding time series

reveals self-similarity or scale-independence. Self-

similarity means that a feature has the same

characteristic value independent of the scale at

which the signal is explored. When the time scale is

changed by a factor m, the statistical distribution

remains unchanged by the factor m

, where H is

called the Hurst scaling exponent. The latter scale-

independence represents the irregularity of the time

series. Fractal methods are amongst those used to

show long-range correlations.

In what follows, in order to focus on the

oscillations of the data, the mean of each signal was

subtracted and the result was divided by the standard

deviation of the original signal before the

computation of the power spectrum.

4.3 Results

From our recordings and results, we first note that

the mean amplitude of LDF signals recorded in the

hand palms is generally higher than the one

observed when the recordings are performed in the

ventral face of the forearms. The latter conclusion

has already been mentioned by other authors (see for

example Freccero et al., 2003). Furthermore, from

the temporal domain analysis, we observe that the

-2

-1

-12

-10

-8

-6

-4

-2

Frequency (Hz)

Power spectrum

Figure 3: Power spectra of LDF signals recorded in a

healthy subject. The red solid curve corresponds to the

power spectrum of the LDF signal recorded in the ventral

face of the forearm; the black dotted line corresponds to

the power spectrum of the LDF signal recorded in the

hand palm.

amplitude variations of LDF signals recorded in the

hands are much higher than in the forearms. Thus, in

average, for the 13 subjects, the amplitude variations

for the hand were 69.5 a.u., whereas they were

13.5 a.u. for the forearm.

An example of power spectrum versus frequency

in logarithmic scales is shown in Figure 3. From all

our power spectrum plots, we observe a clear peak

around 1 Hz for both the hand and the forearm. This

peak is at exactly the same frequency for the signals

recorded simultaneously. This peak is therefore

probably of central origin and probably corresponds

to the heart beat. Furthermore, another peak is

visible around 0.3 Hz, at exactly the same frequency

for the two signals recorded simultaneously. This

peak may be due to the respiration of the subject.

The power spectrum plots also show three

regions, with different slopes. The changes in slopes

occur around 1 Hz and 5 Hz. This is in accordance

with the work of Popivanov et al. (1999). Our results

show that cutaneous LDF signals may exhibit scale-

independence. This has already been predicted by

other authors (see for example Popivanov et al.,

1999). However, the slopes for these three regions

are different for hands and forearms (see an example

in Figure 3). From our knowledge, no comparison

between glabrous and non glabrous skin has already

been published.

Such power spectrum studies have already been

performed on central cardiovascular data (heart rate

variability) and they led to the same conclusion (see

for example Ivanov et al., 2001 and its references).

COMPARISON BETWEEN LASER DOPPLER FLOWMETRY SIGNALS RECORDED IN GLABROUS AND NON

GLABROUS SKIN - Time and Frequency Analyses

137

5 CONCLUSIONS

This study shows that the fluctuations of cutaneous

blood flow (cutaneous LDF signals) recorded in

healthy human subjects are different in

hand (glabrous skin) and forearm (non glabrous

skin). These differences are observed in both the

temporal and the spectral domains. Thus, the mean

amplitude of LDF signals recorded in the hand is

generally higher than in the forearm, and the

fluctuations observed in the hand are much higher

than the ones in the forearm. Furthermore, our work

shows that the power spectrum of LDF signals

recorded in hand and forearm of healthy subjects

may be different. They both may possess

characteristics of fractal processes, but these

characteristics are different for the two analyzed

anatomical sites.

In this paper, a monofractal study has been

performed through the power spectral density

analysis. A multifractal analysis could also be

carried out. Multifractal time series are

heterogeneous, self-similar only in local ranges of

the structure and their fractal measure does vary in

time; hence, they can be characterized by a set of

local fractal measures. Some papers have recently

been published on this field of interest for LDF

data (see for example Humeau et al., 2009; Humeau

et al., 2008).

ACKNOWLEDGEMENTS

The authors thank the “Instituto de Investigação

Interdisciplinar (III)” of the University of Coimbra,

“Acções Universitárias Integradas Luso–Francesas”

(PAUILF) programme and “Fundação para a

Ciência e a Tecnologia (FCT), Lisbon”, for

supporting this work.

REFERENCES

Berardesca E., Lévêque J. L. and Masson P. (2002).

EEMCO Guidance for the Measurement of Skin

Microcirculation. Skin Pharmacology and Applied

Skin Physiology, 15, 442-456.

Fredriksson I., Fors C. and Johansson J. (2007). Laser

Doppler Flowmetry - a Theoretical Framework.

Department of Biomedical Engineering, Linköping

University.

Ivanov P. C., Amaral L. A. N., Goldberger A. L., Havlin

S., Rosenblum M. G., Stanley H. E. and Struzik Z. R.

(2001). From 1/f noise to multifractal cascades in

heartbeat dynamics. Chaos, 11(3), 641-652.

Humeau A., Buard B., Chapeau-Blondeau F., Rousseau

D., Mahe G. and Abraham P. (2009). Multifractal

analysis of central (electrocardiography) and

peripheral (laser Doppler flowmetry) cardiovascular

time series from healthy human subjects.

Physiological Measurement, 30, 617-629.

Humeau A., Chapeau-Blondeau F., Rousseau D.,

Rousseau P., Trzepizur W. and Abraham P. (2008).

Multifractality, sample entropy, and wavelet analyses

for age-related changes in the peripheral

cardiovascular system: Preliminary results. Medical

Physics, 35, 717-723.

Lossius K., Eriksen M. and Walloe L. (1992). Fluctuations

in blood flow to acral skin in humans: connection with

heart rate and blood pressure variability. Journal of

Physiology, 460:641-655.

Morales F. (2005). Improving the clinical applicability of

laser Doppler perfusion monitoring. PhD thesis,

Groningen University.

Nilsson G. E., Salerud E. G., Strömberg T. and Wardell K.

(2003). Laser Doppler perfusion monitoring and

imaging, In: Vo-Dinh T (eds) Biomedical photonics

handbook. CRC Press, Boca Raton.

Popivanov D., Mineva A., Dushanova J., 1999, ‘Dynamic

characteristics of laser-Doppler flux data’, Technology

and Health Care, 7:205-218.

Roustit M., Simmons G. H., Carpentier P. and Cracowski

J. L. (2008). Abnormal digital neurovascular response

to local heating in systemic sclerosis. Rheumatology,

47, 860-864.

Roustit M., Blaise S. and Cracowski J. L. (2009). Sodium

iontophoresis on the finger pad does not consistently

increase skin blood flow in healthy controls and

patients with systemic sclerosis. Microvascular

Research, 77(3), 260-264.

Saumet J. L., Degoute C. S., Saumet M. and Abraham P.

(1992). The effect of nerve blockage on forearm and

finger skin blood flow during body heating and

cooling. International Journal of microcirculation,

11(3), 231-40.

Stefanovska A., Bracic M. and Kvernmo H.D. (1999).

Wavelet analysis of oscillations in the peripheral blood

circulation measured by laser Doppler technique.

IEEE Transactions on Biomedical Engineering, 46,

1230–1239.

Thoresen M. and Walloe T. (1980). Skin blood flow in

humans as a function of environmental temperature

measured by ultrasound. Acta Physiologica

Scandinavica, 109, 333-341.

Tucker A. T., Pearson R. M., Cooke E. D. and Benjamin

N. (1998). Effect of nitric-oxide-generating system

on microcirculatory blood flow in skin of patients with

severe Raynaud’s syndrome: a randomized trial. The

Lancet

, 354, 1670-1675.

Wilson T. E., Zhang R., Levine B. D. and Crandall C. G.

(2005). Dynamic autoregulation of the cutaneous

circulation: differential control in glabrous vs. non-

glabrous skin. American Journal of Physiology, Heart

and Circulatory Physiology, 289, H385-H391.

BIOSIGNALS 2010 - International Conference on Bio-inspired Systems and Signal Processing

138