Optical Non-invasive Flowmetry without Lasers and Coherent Light

D. A. Rogatkin

, D. G. Lapitan

1,2 b

and S. Persheyev

Moscow Regional Research and Clinical Institute “MONIKI” after M.F.Vladimirskiy, 61/2 Shepkina str., Moscow,

Russian Federation

LLC “ODS-MED”, Pushchino, Microdistrict “B”, b.2, Moscow Region, Russian Federation

School of Physics and Astronomy, St Andrews University, St Andrews, U.K.

Keywords: Laser, Flowmetry, Blood Flow, Doppler Effect, Tissue, Spectrum, Intensity, Fluctuation, Incoherent Light.

Abstract: Laser Doppler Flowmetry (LDF) and other optical technique to measure a blood flow in tissues noninvasively

(in vivo) are well-known today. Meanwhile, in spite of more than 40-year history, they do not have

applications in real clinical practice yet. This situation could be a consequence of incorrect understanding of

the physical basics of these methods and, accordingly, of insufficient hardware design, software algorithms,

as well as of erroneous interpretation of the data measured. The basic theory of physical principles of LDF is

the model developed by R.Bonner and R.Nossal in 1980. However, it does not describe many phenomena,

low-frequency fluctuations of optical fields due to a variable blood content in a tissue diagnostic volume, for

example. In this study, we assumed that the low-frequency part of the power spectrum could provide the same

information about the blood flow as the middle- and high-frequency parts provide it in LDF. Moreover, we

proposed the use of coherent light source could be avoided in this case. We have developed a much simpler

and low-cost LED-based prototype and confirmed our assumptions in experiments. Thus, we proposed a new

technique to build simple and economic optical diagnostic tool to evaluate a blood flow in tissues.

1 INTRODUCTION

Optical noninvasive diagnostic techniques, which use

lasers and coherent light for assessment of a tissue

blood flow, such as Laser Doppler Flowmetry (LDF),

Laser Speckle Contrast Imaging (LSCI), etc., are

well-known today. All of them have already proven

its usefulness in a number of medical disciplines

(Briers, 2001), (Rajan et al., 2009), (Roustit et al.,

2012). However, in spite of more than 40-year

history, they are not used yet in a clinical practice

daily. They have many implementations in different

medical research, but their practical applications,

without which a practicing clinician cannot work

today, are not known. Large fluctuations of the output

as well as a low reproducibility of the result often lead

to an inability of the personal diagnostic conclusion

with the use of this technique. Only at scientific

studies in groups of patients, when data are averaged

in groups, there are steadily observed significant

differences in groups.

https://orcid.org/0000-0002-7755-308X

https://orcid.org/0000-0003-3862-0144

https://orcid.org/0000-0003-3970-0488

In our opinion, this situation can be a consequence

of incorrect understanding of the physical basics of

these methods and, accordingly, of not enough correct

hardware design, software algorithms, as well as of

erroneous interpretation of the data measured. If to

consider, for example, the LDF technique, anyone

can see that in the LDF theory the formation of low-

frequency components of the input optical signal is

poorly explained. The basic theory of forming the

optical input signal in LDF is the well-known model

developed by R. Bonner and R. Nossal (B&N model)

(Bonner and Nossal, 1981). Physically based on the

light-beating spectroscopy (Cummins et al., 1970)

and the Doppler effect at light scattering on moving

red blood cells (RBCs) (Nilsson et al., 1980), this

model became the most used and, practically, the

almost single-used theory of LDF. Authors derived

and introduced a power spectrum of the analyzed

signal in the form of the exponential decay, similar to

a fractal noise (1/ω noise, where ω is a frequency).

This power spectrum was then well confirmed in

Rogatkin, D., Lapitan, D. and Persheyev, S.

Optical Non-invasive Flowmetry without Lasers and Coherent Light.

DOI: 10.5220/0009098402150220

In Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 1: BIODEVICES, pages 215-220

ISBN: 978-989-758-398-8; ISSN: 2184-4305

215

experiments (Obeid et al., 1988). However, the nature

of the spectrum can be different, controversial and

debatable, in our opinion. In the B&N model, this

spectrum on a photodetector’s surface is only formed

by the coherent light beating components due to a

heterodyne mixing of the initial probing optical field

and the field having a Doppler frequency shift at light

interaction with moving RBCs inside vessels. Neither

the unsteadiness of scattering properties of tissues

surrounding the vessels due to their compression by

pulsating vessels’ walls, nor any low-frequency

fluctuations (LFFs) of RBCs amount in a diagnostic

volume are not taken into account. Meanwhile, in

recent years a number of authors showed that LFF in

tissue scattering properties can be reflected in the

registered optical signal (Nippolainen et al., 2015),

that variable RBCs content in a diagnostic volume

may play an important role in formation of a low-

frequency part of the power spectrum (Lapitan and

Rogatkin, 2016), or that in LDF the same power

spectrum P(ω) can theoretically be derived from

completely other assumptions (Lapitan et al., 2017).

Therefore, in this study we proposed that the low-

frequency part of the power spectrum could provide

the same information about the blood flow as the

middle- and high-frequency parts of it. Moreover, we

proposed that the coherent light is not mandatory to

form and to detect these LFF. We have developed a

novel, simple, not expensive LED-based prototype

and confirmed our assumptions in experiments.

2 THEORETICAL

BACKGROUND

Due to the LDF and LSCI methods are the most

prevalent at present, as well as due to their results can

be converted into each other so that they can be

considered as of the one family techniques (Bi et al.,

2015), (Fredriksson et al., 2016), we will only

consider in this study the LDF theory, the B&N

model, in particular. The main theoretical statement

of the B&N model, which is also used in all other

versions of the LDF theory (Fredriksson et al., 2007),

(Binzoni and Martelli, 2017), is that the blood

Perfusion Index (PI) or the Blood Flow (BF) can be

determined by analysis of the spectral power density

of the recorded photocurrent (Bonner and Nossal,

1981), (Nilsson et al., 1980). A number of authors use

the amplitude spectrum of the photocurrent or a

photovoltage (Obeid et al., 1988), (Rajan et al., 2009),

but this does not greatly affect the output.

In the case of a photocurrent power spectrum, BF

in LDF is determined by the equation:

(1)

Here ω denotes the angular frequency, P(ω) – the

power spectrum or the amplitude spectrum of a

photocurrent i(t), k

– a dimensional coefficient of a

proportionality. In the case of a power spectrum, P(ω)

can be calculated with the use of the well-known

Wiener–Khintchine theorem (Cummins et al., 1970):













〈









  

〉











(2)

The typical in LDF power spectrum P(ω) of a

photocurrent i(t) is presented in Figure 1.

Figure 1: The typical power spectrum P(ω) of a

photocurrent i(t) described in the B&N model by Eq.2.

In the case of a photocurrent amplitude spectrum,

an equation similar to Eq.1 is used to calculate BF.

The only difference is in the constant k

and in the

normalization parameter <i(t)>

. Normalization for

photocurrent amplitude spectra should be performed

using the constant component of the photocurrent

=<i(t)>. Thus, to evaluate BF in tissues the power

spectrum or the signal amplitude spectrum is only the

key input physical values for data processing. Factors

that form these spectra do not play any important role

for the explored problem.

Usually, in LDF the region of the Doppler effect

ranges from ω

=30 Hz to ω

=30 kHz (Koelink et al.,

1994). Below 30 Hz a photoplethysmographic effect

and motional artifacts are considered dominant.

Therefore, the lower waveband is assumed to be not

useful to calculate BF (Bonner and Nossal, 1981) and,

so, reputedly, not used in LDF-meters by means of

hardware filtering.

Nevertheless, in many publications the spectral

density of a photocurrent or a photovoltage amplitude

spectrum covering the range 1-500 Hz with a

maximum of amplitudes at units or tens of Hz are











〈



〉





∙

















BIODEVICES 2020 - 13th International Conference on Biomedical Electronics and Devices

216

presented (Obeid et al., 1988), (Hu et al., 2013). For

example, Figure 2 represents the

i(t)

spectrum

measured in a portable LDF-meter (Hu et al., 2013).

Figure 2: The low-frequency spectrum of the i(t)

inside a

portable LDF-meter

(Hu et al., 2013).

Recently it was additionally showed (Lapitan et

al., 2018), that in a number of commercially available

LDF-meters, LFFs of the input optical signal can pass

to the output and can influence on the BF calculated

through the normalization parameter (denominator)

<i(t)>

. Usually, this mean value is calculated with

the use of a time window of approximately 0,5-1

second. Therefore <i(t)> can have LFFs below 1-2

Hz. Being squared this will give a wider range of the

spectrum. Generally speaking, LSCI technique is

based on the temporal LFFs of the speckle pattern

contrast or of the field correlation function g

(1)

at a

coherent illumination, as well (Fredriksson et al.,

2016). In LDF, the existence of the optical LFFs in a

tissue microvasculature at external illumination by

stationary-power laser light is now well confirmed

both in experiments (Mizeva et al., 2015), (Mizeva et

al., 2016), and theoretically (Lapitan and Rogatkin,

2016). In the latter case, LFFs were derived as a result

of a variable blood content in the microvasculature.

However, the B&N model does not describe any

LFFs of optical signals. The model was formulated at

the assumption, that amplitudes of all scattered fields

are stationary. In the theory, the blood volume in a

microvasculature is stationary, as well. However,

inside alive tissues these assumptions are wrong.

In our theoretical assumption, we relied on a

number of recently published data mentioned above.

First, we took into account that variable hyperaemia

can form LFFs of the input optical signal (Lapitan and

Rogatkin, 2016). Then, we considered the opinion

and experimental data that vessels’ walls motions at

hart beating compress surrounding connective tissues

changing their optical properties (Nippolainen et al.,

2015). At last, we took into consideration the fact that

the denominator <i(t)> can have LFFs, as well. All

these phenomena form the total LFFs spectra of the

registered and processed signal, which can be used to

calculate BF similar to LDF technique, but with the

use of the low-frequency waveband, below 30 Hz.

Moreover, we assumed, in this case coherent light to

form LFFs and to evaluate BF is not mandatory, so

light emitted diodes (LEDs) can be used as a source

of optical radiation. This technique we named as

Incoherent Optical Fluctuation Flowmetry (IOFF).

3 EXPERIMENTAL PART

3.1 Experimental Prototype

An experimental LED-based prototype that performs

the above method for measuring the skin BF (the skin

blood perfusion) was developed. The appearance of

the prototype is shown in Figure 3.

Figure 3: The appearance of the developed prototype for

measuring the skin blood flow (BF).

The prototype consists of the external optical

probe, the main electronic unit and a laptop with the

special software. In the optical probe, six green-light

LEDs for illuminating the examined skin are placed

radially around a photodetector - a silicon photodiode

– to provide a uniform illumination. The photodiode

registers backscattered radiation from a diagnostic

volume of skin under the optical probe. The narrow-

band radiation in the green spectral range of 560–580

nm was selected as probing radiation to use the

corresponding isosbestic point, at which the light

absorption by oxyhemoglobin and deoxyhemoglobin

in blood is equal. It prevented inaccuracies associated

with different light absorption by venous and arterial

fractions of blood in the diagnostic volume, because

Optical Non-invasive Flowmetry without Lasers and Coherent Light

217

the total signal from skin was collected by the

photodiode regardless of the percentage of

oxyhemoglobin/deoxyhemoglobin concentrations in

the tested volume of skin (diagnostic volume).

Incoherent illumination, LEDs, and the waveband

below 30 Hz allowed us to perform a pulsed regime

of the skin illumination. It is useful to avoid complex

differential scheme of the signal registration, which is

often used in LDF to compensate ambient light and

which is a source of the false spectra formation inside

the instrument (Lapitan et al, 2018). We used the

pulsed switching-on/switching-off regime for LEDs

with 50% duty cycle. The 320 Hz operating frequency

was chosen to satisfy the Nyquist criterion with the 5-

fold margin. During the switching-on time of LEDs

total backscattered probing radiation together with

existing ambient light were registered, while during

the switching-off time of LEDs the backscattered

ambient light was only registered and analysed.

Subtracting the ambient light signal from the total

signal registered, we eliminated the ambient radiation

impact on the registered probing radiation.

The purified from ambient light signal was further

amplified, digitized at the frequency of 320 Hz, and

processed in a laptop by the LabVieW-based software

to compute BF (Figure 4).

Figure 4: The block-diagram of data processing. BPF is the

band-pass filter, DAB is the data accumulation buffer, FFT

is the Fast Fourier Transform module, ∑ is the adder, DIV

is the divider, DC is the unit for extraction of the <i(t)>.

We used the photocurrent i(t) amplitude spectrum

technique to compute BF. First, the ac component of

the signal is extracted by a band-pass numerical

Butterworth filter (BPF) of the 2nd order (see the

upper way of the block-diagram). The frequency

range was chosen as follows: ω

=0,5 Hz and ω

=12

Hz. Next, the ac signal is accumulated in the data

buffer (DAB) within each 1 second. For this purpose,

a buffer sample size of 320 points was used. These

320 points are then directed every second to the Fast

Fourier Transform (FFT) module, where the

amplitude spectrum of the signal is formed. The

resolution of such a spectrum is 1 Hz. Finally, all

spectral components are summed in the adder Σ with

corresponding frequency weights (numerical

integration) and divided by the dc component of

i(t).

The dc component is obtained by averaging of all 320

signal magnitudes on the time interval 1 second (see

the lower way of the block-diagram). Thus, BF points

are formed as the output every 1 second.

To perform functional tests with skin heating, a

heating metal plate was incorporated in the optical

probe. Heating was performed by a pulsed current

with a pulse-width automatic modulation. An

operator can set the desired heating temperature of the

plate with a given heating rate.

3.2 Experimental Study and Results

At the initial step of our experimental study, to

confirm the presence in the spectra the proposed LFFs

at a continuous incoherent illumination of the tested

skin, we measured a photocurrent power spectrum

after the photodiode with the use of the standard

spectral equipment (L-CARD spectral analyzer, RF).

Figure 5 represents the typical spectral power density

of i(t) at the green-light LEDs illumination of the

fingerprint skin area of a volunteer.

Figure 5: Typical spectral power density of i(t) in in vivo

experiments. 1 - Healthy volunteer at a rest. 2 - Healthy

volunteer at the shoulder arterial occlusion.

As one can see, the normal blood flow in arms

forms visible LFFs in the spectral range 0-12 Hz,

while the arterial occlusion blocks LFFs. It confirms

well our assumptions. Doppler components in i(t) are

not presented in these spectra due to the absence of

the coherent illumination. It is also interesting to note

that the blood-pulsed spectral components with the

frequency of around 1 Hz are a visible part of the total

LFFs spectrum (see the spectrum 1), but they do not

form the power spectrum completely, as it might

seem from the theory of a photoplethysmography.

In addition, at the arterial occlusion we measured

and analyzed a behavior of the photovoltage dc

component. Like for a photocurrent, the photovoltage

dc component was calculated by averaging all 320

digitized signal points on a time interval of each one

second. Figure 6 shows the example of the behavior

and a corresponding BF computed with the use of this

photovoltage U. The example clearly shows a

decrease in ac components of the signal (decrease in

LFFs), and an increment dc of the dc component of

U at arterial occlusions.

BIODEVICES 2020 - 13th International Conference on Biomedical Electronics and Devices

218

Figure 6: Example of the recorded dc photovoltage U

during a test with the shoulder arterial occlusion and the

corresponding BF computed.

Final steps of this our experimental study were

comparative measurements of BF with the use of our

new method and a standard LDF technique. Standard

LDF-meter LAKK-02 (LAZMA Ltd., RF) working at

the isosbestic point 808 nm was used as the reference

tool. Functional tests with skin heating and arterial

occlusion were carried out for the comparative study.

Figure 7 shows the design of the study. BF was

simultaneously measured in the outer side of a

forearm by IOFF optical probe (1) and by the standard

LDF optical fiber probe (2). To execute a test with

skin heating under the fiber optical probe, an external

heating probe was used. Final heating temperature for

both probes was 42

C. The same arm was explored to

execute tests with occlusion. Occlusion pressure was

applied by a standard tonometer’s cuff inflation.

Figure 7: Design of the comparative study.

To made results of two different measurements

comparable in BF magnitudes (in pf. un. - perfusion

units), it was necessary to select properly the

proportionality coefficient k

for the IOFF technique

in the main Equation 1, because for the standard LDF-

meter it was already embedded in its software and not

changeable. The needed value of the coefficient k

for

new IOFF technique was obtained by means of the

selection of the approximate equality of BF

magnitudes during the first test with occlusion.

Examples of the recorded BFs in these final

experiments are shown in Figure. 8.

Figure 8: Examples of the recorded BF during heating (top)

and arterial occlusion (bottom) tests with the simultaneous

usage of IOFF and LDF techniques.

As seen in the Figure 8, the measured BFs were

similar even in details at arterial occlusions, while at

the heating test IOFF technique showed the enhanced

sensitivity. Probably, it is a consequence of different

wavebands used – green in the IOFF case and near

infrared in the LDF one. Nevertheless, the result is

visible. Our assumptions were confirmed well.

4 CONCLUSIONS

Relying on a number of recently published data, in

this study we assumed a strong influence of low-

frequency fluctuations (LFFs) in registered optical

signals on the final output in LDF. LFFs form the

low-frequency spectra of all processed signals, and

can be used to calculate BF similar to LDF technique,

but using the waveband below 30 Hz. Moreover, we

assumed that the coherent illumination and lasers to

evaluate BF is not mandatory in this case. This

technique we named as IOFF - Incoherent Optical

Fluctuation Flowmetry. We developed a LED-based

prototype that performs IOFF, and carried out a

number of experiments to confirm our assumptions.

Optical Non-invasive Flowmetry without Lasers and Coherent Light

219

Most of our experiments confirmed assumptions

we made well. LFFs were registered with incoherent

illumination in skin. The main spectral range of LFFs

was determined between 0 and 12 Hz. It allowed us

to calculate BF similar to the LDF algorithm, but

inside the waveband below 30 Hz. Comparative

measurements of BF using our novel method and a

standard LDF technique showed a good similarity of

the results. Measured BFs were equal even in details

at arterial occlusions, while at heating tests IOFF

technique showed the enhanced sensitivity. These

positive results open a way for building novel and less

sophisticated than LDF optical diagnostic tools for

assessment of BF in tissues. Of course, the proposed

IOFF technique needs further detailed investigations,

especially in clinics to prove its clinical significance.

However, as one can see, our approach already has a

number of additional advantages. One important

advantage is the cost of the equipment. A commercial

LDF-meter such as the Moor VMS-LDF costs more

than 10,000 USD. The cost of our self-designed

portable prototype is less than 100 USD (including all

components except a computer). The second one is

not sophisticated and clear metrology. The metrology

in LDF is sophisticated due to a complexity with the

design of tissue-like phantoms imitating the motion

of RBCs in a microvasculature bed. In our case, an

imitation of the amplitude modulation of the probing

radiation with different modulation depths on the

background of different levels of the dc component of

the backscattered radiation is sufficient.

REFERENCES

Bi, R., Dong, J., Poh, C.L., and Lee, K., 2015. Optical

methods for blood perfusion measurement – theoretical

comparison among four different modalities. J. Opt.

Soc. Am. A., 32(5), 860- 866.

Binzoni, T., and Martelli, F., 2017. Study on the

mathematical relationship existing between single-

photon laser-Doppler flowmetry and diffuse correlation

spectroscopy with static background. J. Opt. Soc. Am.

A., 34(12), 2096-2101.

Bonner, R. and Nossal, R., 1981. Model for laser Doppler

measurement of blood flow in tissue. Applied Optics,

20(12), 2097-2107.

Briers, J.D., 2001. Laser Doppler, speckle and related

techniques for blood perfusion mapping and imaging.

Physiological measurement, 22(4), R35-R66.

Cummins, H.Z. and Swinney, H.L., 1970. Light Beating

Spectroscopy. Progress in Optics, 8, 133-200.

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

Doppler Flowmetry – a Theoretical Framework.

Department of Biomed. Engineering, Linköping Univ.

Fredriksson, I., Larsson M., 2016. On the equivalence and

differences between laser Doppler flowmetry and laser

speckle contrast analysis. J. of Biomed. Optics, 21(12),

126018.

Hu, C.L., Lin, Z.S., Chen, Y.Y., Lin, Y.H., Li, M.L., 2013.

Portable laser Doppler flowmeter for microcirculation

detection. Biomed. Engineering Letters, 3(2), 109-114.

Koelink, M.H., De Mul, F.F.M., Leerkotte, B., et al., 1994.

Signal processing for a laser-Doppler blood perfusion

meter. Signal processing, 38(2), 239-252.

Lapitan, D.G., Rogatkin, D.A., 2016. Variable hyperemia

of biological tissue as a noise source in the input optical

signal of a medical laser Doppler flowmeter. J. Opt.

Techn., 83(1), 36-42.

Lapitan, D., Rogatkin, D., Persheyev, S. and Kotliar, K.,

2018. False spectra formation in the differential two-

channel scheme of the laser Doppler flowmeter.

Biomed. Eng.-Biomed. Tech., 63(4), 439-444.

Lapitan, D., Rogatkin, D., Persheyev, S. and Rogatkin, A.,

2017. New simple phenomenological model for laser

Doppler measurements of blood flow in tissue. Proc. of

the 10th Int. Joint Conference on Biomedical

Engineering Systems and Technologies (BIOSTEC

2017), vol.1: BIODEVICES, 98-103.

Mizeva, I., Maria, C., Frick, P., Podtaev, S., Allen, J., 2015.

Quantifying the correlation between

photoplethysmography and laser Doppler flowmetry

microvascular low-frequency oscillations. J. of Biomed.

Optics, 20(3), 037007.

Mizeva, I., Frick, P., Podtaev, S., 2016. Relationship of

oscillating and average components of laser Doppler

flowmetry signal. J. of Biomed. Optics, 21(8), 085002.

Nilsson, G.E., Tenland, T., Oberg, P.A., 1980. A new

instrument for continuous measurement of tissue blood

flow by light beating spectroscopy. IEEE Transactions

on Biomed. Engineering, 27(1), 12-19.

Nippolainen, E., Podolian, N.P., Romashko, R.V., Kulchin

Y.N., Kamshilin A.A., 2015. Photoplethysmographic

wavwform as a function of subject’s age. Physics

Procedia, 73, 241-245.

Obeid, A.N., Boggett, D.M., Barnett, N.J., Dougherty, G.,

and Rolfe, P., 1988. Depth discrimination in laser

Doppler skin blood flow measurement using different

lasers, Med. & Biol. Engin. & Comput., 26, 415-419.

Rajan, V., Varghese, B., Leeuwen, T., 2009. Review of

methodological developments in laser Doppler

flowmetry. Lasers Med Sci, 24, 269–283.

Roustit, M., Cracowski, J., 2012. Non-invasive assessment

of skin microvascular function in humans: an insight

into methods. Microcirculation, 19(1), 47-64.

BIODEVICES 2020 - 13th International Conference on Biomedical Electronics and Devices

220