Application of a MEMS Blood Flowmeter for Power Spectrum

Analysis of Heart Rate Variability

Terukazu Akiyama

, Tatsuya Miyazaki

, Hiroki Ito

, Hirofumi Nogami

and Renshi Sawada

1,2

Graduate School of Systems Life Sciences, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka-city, Japan

Department of Mechanical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka-city, Japan

Keywords: MEMS Blood Flowmeter, Heart Rate Variability, Autonomic Nervous System.

Abstract: We investigated the possibility of applying a MEMS blood flowmeter to heart rate variability (HRV)

analysis. We conducted simultaneous measurements of HRV by electrocardiogram and MEMS blood

flowmeter. TPP for the MEMS blood flowmeter was defined as the interval between peaks, which were

designated as where the first-order differential of the signal changes from negative to positive. TRR (i.e., the

R-R interval of the electrocardiogram) and TPP were compared by regression analysis. Autonomic indices

transformed by power spectrum analysis were also compared by regression analysis. Fast Fourier transform

(FFT) and maximum entropy method (MEM) were employed in the frequency analysis. By FFT analysis,

the coefficient of determination for the regression between LF%, HF%, and LF/HF derived by TRR versus

TPP was 0.8781, 0.8781, and 0.8946, respectively. By MEM analysis, the coefficient of determination for

the regression between LF%, HF%, and LF/HF derived by TRR versus TPP was 0.9649, 0.8026, and

0.9181, respectively. These high correlations suggest that the TPP of the MEMS blood flowmeter is a

reliable metric that can be utilized in applications of HRV analysis.

1 INTRODUCTION

The autonomic nervous system plays key roles

throughout the body, including the control of

involuntary activities such as blood circulation,

respiration, digestion, regulation of body

temperature through perspiration, endocrine

function, generative function, and metabolism. One

noteworthy example is the pumping of the heart,

which is caused by periodic excitement of the

sinoatrial node, a rhythm that typically fluctuates.

Fluctuation of heartbeat rhythm is known as heart

rate variability (HRV), and it reflects the health of

the cardiovascular autonomic nervous system.

Research into HRV began in dogs in 1981 using

quantitative spectral analysis with the FFT, and it

has since been studied in humans (Akselrod et al.,

1981, Pomeranz et al., 1985). The prospect that

HRV might be measurable more easily without the

need for large equipment attracted much attention

within the information engineering and biomedical

engineering fields in the 1990s, with advances in

wireless networks and miniaturization of biological

sensors at that time informing attitudes that

considered the prospect feasible. As the autonomic

nervous system innervates many organs, symptoms

of its dysfunction are observed in various diseases,

and recordings of parameters it controls may be

useful for studying the occurrence, pathogenesis,

response to treatment, and prognosis of a multitude

of diseases. While many clinical applications of

HRV analysis have been reported, it has been most

accurately used as a clinical marker of diabetic

neuropathy, as well as a predictive factor for the risk

of sudden cardiac death and the occurrence of

arrhythmias after acute myocardial infarction (Stys

et al., 1998). Moreover, it has been employed to

identify neurological diseases and to confirm the

effects of therapy and medicine targeting them.

Because of the wide range of disease factors

(pathology, prognosis, etc.) that are related to

autonomic nervous system function and dysfunction,

it is a highly regarded health metric in many fields.

However, symptoms of its dysfunction are more

difficult to capture via short-term measurements

compared with other health indices; indeed, in cases

where its function fluctuates based on circadian

changes (e.g., temperature regulation, where body

temperature is cooler during the nighttime), accurate

evaluation becomes difficult because of the need to

211

Akiyama T., Miyazaki T., Ito H., Nogami H. and Sawada R..

Application of a MEMS Blood Flowmeter for Power Spectrum Analysis of Heart Rate Variability.

DOI: 10.5220/0005213302110218

In Proceedings of the International Conference on Bio-inspired Systems and Signal Processing (BIOSIGNALS-2015), pages 211-218

ISBN: 978-989-758-069-7

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

collect data through continuous monitoring

throughout a subject’s daily activities. Autonomic

nervous system function reacts to situation and

environment in real-time to maintain homeostasis.

Early detection of symptoms of its dysfunction may

be possible by analyzing reactions to return the body

to homeostasis. For example, in the case of diabetes,

the body would be slow to release insulin in

response to high blood sugar; pancreatic

neurological dysfunction could be reflected in HRV.

If these data were accessible by medical institutions

that provide health and medical services, better and

more-effective service could likely be provided.

To this end, we have recently developed a

perhaps more widely applicable solution: a micro-

integrated laser Doppler blood flowmeter (micro-

electromechanical system (MEMS) blood flow

sensor), which is 1/300

the volume of a

conventional laser Doppler blood flowmeter

(Higurashi et al., 2003, Kimura et al., 2010). The

weight of this sensor is approximately 1/30

of a

conventional instrument, and its power consumption

is only 1/20

. The instrument also makes blood flow

measurements possible while subjects are moving by

omitting the optical fiber that is the source of

significant noise in conventional instruments. This

makes MEMS blood flowmetry ideal for the

continuous measurement of blood flow volume over

time. Moreover, since the blood flow data obtained

by MEMS blood flowmetry are a result of analyzing

the skin sympathetic nerves, blood pressure, and

body temperature, MEMS blood flowmetry is

capable of performing multivariate analysis of many

indices, a significant advantage over single-index

methods in the efforts to understand the complex

systems of the living body.

The R-wave of the electrocardiogram is one

candidate parameter often used in HRV analysis to

assess autonomic function. Several commercial

products on the market today allow ECG

measurements during training/movement with an

acceptable accuracy. However, we believe that a

MEMS blood flowmeter is advantageous for

evaluation of the autonomic nervous system for two

reasons. One reason is that a MEMS blood

flowmeter is a small and simple system and, unlike

the ECG, does not require operator skill to affix it in

order to obtain pulse. Another reason is its ready

applicability to multivariable analysis for the

evaluation of autonomic nervous function. If

autonomic nervous system-based changes in pulse

obtained from the pulse waveform of blood flow

volume can be accurately determined, both simple

measurement and further detailed evaluation of

biological state would be possible, which would be a

boon to applications in healthcare and medical

treatments.

At present, R-wave (obtained from

electrocardiography) and pulse (obtained from

acceleration plethysmography) data are typically

used in HRV analysis (Takada et al., 2008). To

confirm whether HRV analysis is possible from the

pulse data obtained from the blood flow volume

waveform of MEMS blood flowmetry, we

performed simultaneous measurement of the R-R

interval (TRR) obtained from an electrocardiogram

and the pulse wave interval (TPP) obtained from a

MEMS blood flowmeter for 100 beats, and then

investigated the relationship between TRR and TPP

by regression analysis (Fig. 1). Furthermore, we

compared the accuracy of the frequency analysis of

the fast Fourier transform (FFT) versus the

maximum entropy method (MEM) for the HRV

signal over 100 beats.

Figure 1: (a) Photograph of the MEMS blood flowmeter;

(b) Measurement principle and schematics of its optical

MEMS chip.

PD: photodiode, DFB-LD: distributed feedback-laser

diode.

BIOSIGNALS2015-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

212

2 MATERIALS AND METHODS

In this section, we provide information on the study

subjects, MEMS blood flowmeter, data analysis, and

experimental method.

2.1 Subjects

The investigation complied with the ethical

standards of the committee responsible for human

experimentation of Kyushu University and with the

Helsinki Declaration of 1964 as revised in 2008. A

total of 6 healthy volunteers (6 males, age range:

21–26 years) participated in this study.

2.2 MEMS Blood Flowmeter

The MEMS blood flowmeter employed in the study

is a blood flowmeter that is integrated with MEMS

technology. Its features of ultra-compactness,

lightweight, low power consumption, and wireless

function make it possible to always be attached to

the body. In addition, since its design omits the

optical fiber that causes significant noise in

conventional blood flowmeters, peripheral blood

flow is measurable while the subject is in motion.

2.2.1 Structure

Figure 1 shows a schematic of the MEMS blood

flowmeter. The MEMS blood flowmeter consists of

the main body and the probe. The main body has a

digital signal processor (DSP), a Bluetooth wireless

link, a display, and a battery. The probe has a sensor

chip and an amplifier circuit. The sensor chip

consists of two crystal silicon substrates, a DFB-LD

(distributed feedback-laser diode) and a PD

(photodiode). One of the silicon substrates contains

two cavities that hold the bonded LD and PD, while

the other contains a microlens and a hole. Gold

etched into the cavity leads the laser beam from the

LD into the microlens. The antireflection-coated

microlens optimizes the laser beam from the LD.

Another silicon substrate is used to seal the cavity.

2.2.2 Principle

The design of the MEMS laser Doppler blood

flowmeter is based on the principle of a laser

Doppler flowmeter (Bonner et al., 1981). The

measurement object is irradiated with a laser beam.

Incident light is scattered multiply by static tissue,

such as a skin and blood vessels, and moving tissue,

such as red blood cells. All light scattered back

Figure 2: Data analysis process.

toward the PD interferes on the PD to form a speckle

pattern. Statistical analysis of the output signal of the

PD with respect to the speckle pattern gives a

relative blood flow.

2.3 Data Analysis

Figure 2 is a flowchart of the data analysis process.

The signal obtained by the MEMS blood flowmeter

is detrended, and smoothing is performed by the

moving average and first-order differential. We

obtain the pulse by detecting the peak of the signal.

The pulse is smoothed by the moving average and

Figure 3: Waveforms of the electrocardiogram and the

MEMS blood flowmeter.

ApplicationofaMEMSBloodFlowmeterforPowerSpectrumAnalysisofHeartRateVariability

213

third-order spline interpolation. Frequency analysis

by the FFT or MEM is performed to obtain the

power spectrum of HRV. Finally, we derive

autonomic indices from the HRV spectrum (details

of index calculations below).

2.3.1 Peak Detection

The signal of the laser Doppler flowmeter is

influenced by peripheral blood flow, which is in turn

influenced by many factors such as blood vessel

conditions, body temperature, vasoconstriction, and

vasodilation by skin sympathetic nerves, vascular

endothelial cell metabolism, respiration, and blood

pressure. In addition, changes in peripheral blood

flow due to heartbeat are smaller than arterial blood

flow. Therefore, we anticipated that the detection of

the signal peak would be difficult compared with

analyses of the clear waveforms in

electrocardiograms and plethysmographs. Methods

of detecting the peak of the pulse waveform may

define the peak as the minimum, maximum, or point

of 10% of the minimum within the waveform. There

are also peak detection methods that instead

emphasize change in the signal in terms of the first

order-differential and second-order differential.

When the latter strategy is applied to peripheral

blood flow, the first-order differential and the

second-order differential may indeed be suitable for

functioning as a high-pass filter for detecting

changes in the signal trend. However, we assumed

that peaks could not be detected more stably with

reference to other, higher-frequency signals in

peripheral blood flow more reliably than simply with

reference to the zero order peak. Therefore, we

consider that the optimum pre-filtering strategy is to

remove only high-frequency fluctuations before the

first-order differential and the second-order

differential. By the above logic, in this study we

decided to perform a first-order differential after

smoothing by the moving average, and detected

peaks as when the signal changed from negative to

positive (Figure 3).

2.3.2 Heart Rate Variability Analysis

Heart rate is controlled by the electrical impulses of

the sinoatrial node, which are in turn controlled by

the central nervous system through the sympathetic

and parasympathetic branches of the autonomic

nervous system. Sympathetic nerves increase heart

rate by accelerating the electrical impulses of the

sinoatrial node, while parasympathetic nerves

decrease heart rate by suppressing them. The central

nervous system also controls various functions based

on the information sent from the sensory organs.

Therefore, fluctuations that are generated by the

body modulate heart rate through the autonomic

nervous system via the central nervous system. Of

particular relevance to our study is that fluctuations

in blood pressure—i.e., the Mayer wave, over a 10-

second period—and respiration cycle are transmitted

by this route to modulate heart rate. An important

fact that can be leveraged for analysis is that the

frequency responses of sympathetic nerves and

parasympathetic nerves are different: about 0.15 and

1 Hz, respectively. As a consequence, although

sympathetic and parasympathetic nerves work

antagonistically in general, heart rate is modulated

by fluctuations in blood pressure (i.e., the 10-s

Mayer wave) when it receives input from the former,

but by fluctuations in both blood pressure and

respiration when it receives input from the latter

(Penaz, 1978). Therefore, it is possible to evaluate

the relative strength of sympathetic and

parasympathetic function on heart rate by frequency

analysis of heart rate fluctuation. These two indices

are defined as follows from each interval:

 Low-frequency component (LF) : 0.04-0.15

Hz of power spectral density

 High-frequency component (HF) : 0.15-0.40

Hz of power spectral density

Thus, it is possible to evaluate the relative

sympathetic function by calculating LF/HF. In

addition, the HF estimates the parasympathetic

function. However, since the HF is a relative value,

it must be normalized by the total power to be of

analytical use.

2.4 Experimental Method

In order to compare the TPP obtained by MEMS

blood flow sensor and TRR obtained by

electrocardiogram, simultaneous measurement with

conventional electrocardiogram and MEMS blood

flowmeter was conducted on six healthy volunteers

(6 males). The experimental protocol is based on the

method of measuring blood flow as follows: after

attaching the conventional electrocardiogram and

MEMS blood flowmeter to the subject’s body, we

provided a break-time of 10 minutes, after which

heart rate was continuously measured for 100 beats.

We conducted experimental measurements in the

sitting position throughout the experiment. The

MEMS blood flowmeter was attached to the right

index finger. The right hand was fixed at the vertical

level of the heart to prevent it from moving.

Measurements of peripheral blood flow are

deformed by changing the measurement site because

BIOSIGNALS2015-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

214

Figure 4: Relationship between the TRR of the electrocardiogram and the TPP of the MEMS blood flowmeter by regression

analysis.

TRR: R-R interval obtained from electrocardiogram, TPP: pulse wave interval obtained from MEMS blood flowmeter.

the density and shape of blood vessels differ across

measurement sites. Therefore, the experiment was

conducted with the flowmeter attached to the same

site throughout. In addition, since contact pressure

by the flowmeter compresses blood vessels and thus

changes peripheral blood flow, the MEMS blood

flowmeter was lightly fixed with double-sided tape.

Furthermore, since peripheral blood flow is altered

by skin temperature, room temperature was held at

constant 25 °C throughout the experiment. The LF

and HF components of the TRR and TPP spectra

were defined as 0.04–0.15 Hz and 0.15–0.40 Hz of

power spectral density, respectively. From the power

spectral density obtained from TRR and TPP, LF%

and HF% (normalized to total power) in addition to

LF/HF were calculated.

3 RESULTS

Figure 4 shows the correlations between TRR and

TPP for each subject. The regression equation and

the coefficient of determination obtained by

regression analysis are also shown. For subjects

ID1–6, the slope was 0.8461, 0.9775, 0.9543,

0.9193, 0.9814, and 0.9554, and the intercept was

ApplicationofaMEMSBloodFlowmeterforPowerSpectrumAnalysisofHeartRateVariability

215

Figure 5: Relationships of LF%, HF%, and LF/HF as calculated by TRR (electrocardiogram) versus by TPP (MEMS blood

flowmeter) by regression analysis. Each dot in the graphs represents a data point from one individual. A total of 6 healthy

male volunteers (age range: 21–26 years) participated in this study. FFT frequency analysis was performed for TRR data

(left side); MEM frequency analysis was performed for TPP data (right side).

LF%: proportion of power spectral density in the low-frequency domain of HRV, HF%: proportion of power spectral

density in the high-frequency domain of HRV, LF/HF: ratio of LF and HF power spectral density, FFT: fast Fourier

transform, MEM: maximum entropy method.

147.16, 17.706, 37.435, 66.507, 17.097, and 32.909,

respectively, with the corresponding coefficients of

determination being 0.6967, 0.8593, 0.8673, 0.831,

0.7794, and 0.65 (p<0.01).

Figure 5 shows the relationships between LF%,

HF%, and LF/HF, as determined by frequency

analysis using FFT versus MEM to analyze TRR

and TPP (respectively). The coefficients of

determination and the regression equation obtained

by regression analysis are also shown.

According to FFT analysis, the slope between

TRR and TPP for LF%, HF%, and LF/HF was

0.7053, 0.7053, and 0.2753, respectively; the

coefficient of intercept was 0.15, 0.1447, and

1.0866. The coefficient of determination was

0.8781, 0.8781, and 0.8946, respectively.

According to MEM analysis, the slope between

TRR and TPP for LF%, HF%, and LF/HF was

BIOSIGNALS2015-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

216

0.8019, 0.7933, and 0.2827, respectively; the

coefficient of intercept was 115975, 172243, and

0.9977. The coefficient of determination was

0.9649, 0.8026, and 0.9181, respectively.

4 DISCUSSION

We conducted a comparison of the RR interval

obtained by an electrocardiogram and the pulse

interval obtained from the pulse signal of our

MEMS blood flowmeter. Coefficients of

determination showed a strong relationship between

the two, and ranged from 0.65–0.87 in the regression

analysis. In addition, autonomic indices obtained via

the former metric showed a strong relationship with

autonomic indices obtained via the latter, with

coefficients of determination ranging from 0.80–

0.96. The MEMS blood flowmeter measures the

pulse signal transmitted from the heart to peripheral

blood vessels. Many factors modulate this pulse

signal. For example, blood vessel structure changes

due to low-frequency fluctuation of skin sympathetic

nerves and vascular endothelial cells. These skin

sympathetic nerves also cause short-term expansion

and contraction of blood vessels. Strong correlation

of blood flow signal modulation with the

electrocardiogram TRR metric was also observed in

this experiment. These results suggest high promise

for evaluating autonomic function and heart rate

variability from the pulse interval of the pulse signal

of MEMS blood flowmeter.

In the future, we will determine whether the

MEMS blood flowmeter can measure relative

changes in autonomic indices induced by

physiological or psychological stressors.

5 CONCLUSIONS

The feasibility of using a MEMS blood flowmeter to

evaluate autonomic function by heart rate variability

was evidenced by comparable findings with

electrocardiogram and pulse waveform data.

Frequency analysis of each signal fluctuation of 100

consecutive heartbeats by the FFT and MEM was

performed, followed by comparison of the resultant

autonomic nervous indices by regression analysis.

We confirmed a strong relationship between

MEMS- and electrocardiogram-derived indices from

a comparison of autonomic function indices and

pulse. We also confirmed the strong relationship

between the electrocardiogram and MEMS blood

flowmeter with power spectrum analysis by both

FFT and MEM. Although MEM has a frequency

resolution higher than the FFT, the FFT also has

comparable ability to analyze data at least to the

extent of the 100 consecutive heartbeats monitored

in this study. The signal obtained by the MEMS

blood flowmeter reflects many indicators such as

skin sympathetic nerve activity (Söderström et al.,

2003), body temperature, and blood pressure, and, to

date, was successfully applied to the evaluation of

hemodynamics during exercise and the detection of

signs of dehydration (Nogami et al., 2011) and

alcohol ingestion (Iwasaki et al., 2012). As

autonomic function evaluation by the MEMS blood

flowmeter becomes more and more feasible in the

future, its applications are expected to grow in a

wide range of fields.

ACKNOWLEDGEMENTS

We would like to express out our special thanks to

the volunteers who participated in the study.

REFERENCES

Akselrod, S., Gordon, D., Ubel, F. A., Shannon, D. C.,

Berger, A. C., & Cohen, R. J. (1981). Power spectrum

analysis of heart rate fluctuation: a quantitative probe

of beat-to-beat cardiovascular control. science,

213(4504), 220-222.

Pomeranz, B., Macaulay, R. J., Caudill, M. A., Kutz, I.,

Adam, D., Gordon, D. A. V. I. D., ... & Cohen, R. J.

(1985). Assessment of autonomic function in humans

by heart rate spectral analysis. American Journal of

Physiology-Heart and Circulatory Physiology, 248(1),

H151-H153.

Stys, A., & Stys, T. (1998). Current clinical applications

of heart rate variability. Clinical cardiology, 21(10),

719-724.

Higurashi, E., Sawada, R., & Ito, T. (2003). An integrated

laser blood flowmeter. Journal of lightwave

technology, 21(3), 591.

Kimura, Y., Goma, M., Onoe, A., Higurashi, E., &

Sawada, R. (2010). Integrated laser Doppler blood

flowmeter designed to enable wafer-level packaging.

Biomedical Engineering, IEEE Transactions on,

57(8), 2026-2033.

Takada, M., Ebara, T., & Sakai, Y. (2008). The

acceleration plethysmography system as a new

physiological technology for evaluating autonomic

modulations. Health evaluation and promotion, 35(4),

373-377.

ApplicationofaMEMSBloodFlowmeterforPowerSpectrumAnalysisofHeartRateVariability

217

Bonner, R., & Nossal, R. (1981). Model for laser Doppler

measurements of blood flow in tissue. Applied optics,

20(12), 2097-2107.

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

doppler flowmetry-a theoretical framework.

Department of Biomedical Engineering, Linköping

University.

Rowell, L. B. (1986). Human circulation: regulation

during physical stress (p. 215). Oxford: Oxford

University Press.

Smith, J. J. (1990). Circulatory response to the upright

posture (No. 6). CRC Press.

Yamamoto, Y., & Hughson, R. L. (1991). Coarse-graining

spectral analysis: for studying heart rate variability.

Penaz, J. (1978). Mayer waves: history and methodology.

Automedica, 2(3), 135-142.

Söderström, T., Stefanovska, A., Veber, M., & Svensson,

H. (2003). Involvement of sympathetic nerve activity

in skin blood flow oscillations in humans. American

Journal of Physiology-Heart and Circulatory

Physiology, 284(5), H1638-H1646.

Nogami, H., Iwasaki, W., Abe, T., Kimura, Y., Onoe, A.,

Higurashi, E., ... & Sawada, R. (2011). Use of a simple

arm-raising test with a portable laser Doppler blood

flow meter to detect dehydration. Proceedings of the

Institution of Mechanical Engineers, Part H: Journal

of Engineering in Medicine, 225(4), 411-419.

Iwasaki, W., Nogami, H., Ito, H., Gotanda, T., Peng, Y.,

Takeuchi, S., ... & Sawada, R. (2012). Useful method

to monitor the physiological effects of alcohol

ingestion by combination of micro-integrated laser

Doppler blood flow meter and arm-raising test.

Proceedings of the Institution of Mechanical

Engineers, Part H: Journal of Engineering in

Medicine, 226(10), 759-765.

Ogata, H., Fujimaru, I., Yamada, K., & Kondo, T. (2012).

Suppression of cardiocirculatory responses to

orthostatic stress by passive walking-like leg

movement in healthy young men. J Physiol Anthropol,

31, 24.

Hess, W. R. (1954). Diencephalon, autonomic and

extrapyramidal functions (Vol. 3). Grune & Stratton.

Ohhashi, T., Sakaguchi, M., & Tsuda, T. (1998). Human

perspiration measurement. Physiological

measurement, 19(4), 449.

Homma, S., Nakajima, Y., Toma, S., Ito, T., & Shibata, T.

(1998). Intracerebral source localization of mental

process-related potentials elicited prior to mental

sweating response in humans. Neuroscience letters,

247(1), 25-28.

Tanaka, H., Yamaguchi, H., Matushima, R., & Tamai, H.

(1999). Instantaneous orthostatic hypotension in

children and adolescents: a new entity of orthostatic

intolerance. Pediatric research, 46

(6), 691-691.

Low, P. A., Opfer-Gehrking, T. L., Textor, S. C.,

Benarroch, E. E., Shen, W. K., Schondorf, R., ... &

Rummans, T. A. (1995). Postural tachycardia

syndrome (POTS). Neurology, 45(4 Suppl 5), S19-25.

Wheeler, T., & Watkins, P. J. (1973). Cardiac denervation

in diabetes. British Medical Journal, 4(5892), 584.

Oka, H., Mochio, S., Sato, K., Sato, H., Katayama, K.,

Watanabe, S., ... & Isogai, Y. (1995). Spectral

analyses of RR interval and systolic blood pressure in

diabetic autonomic neuropathy. Journal of the

autonomic nervous system, 52(2), 203-211.

Bonner, R. F., Clem, T. R., Bowen, P. D., & Bowman, R.

L. (1981). Laser-Doppler continuous real-time monitor

of pulsatile and mean blood flow in tissue

microcirculation. In Scattering Techniques Applied to

Supramolecular and Nonequilibrium Systems (pp.

685-701). Springer US.

Kano, T., Shimoda, O., Higashi, K., Sadanaga, M., &

Sakamoto, M. (1993). Fundamental patterns and

characteristics of the laser-Doppler skin blood flow

waves recorded from the finger or toe. Journal of the

autonomic nervous system, 45(3), 191-199.

Abe, M., Ando, Y., Higashi, K., & Kano, T. (1996). Non-

neurogenic periodic fluctuations in heart rate and

vasomotion appearing in familial amyloid

polyneuropathy (FAP) Type I (Met30). Journal of the

autonomic nervous system, 60(1), 71-75.

Ide, J., Yamaga, M., Kitamura, T., & Takagi, K. (1997).

Quantitative evaluation of sympathetic nervous system

dysfunction in patients with reflex sympathetic

dystrophy. The Journal of Hand Surgery: British &

European Volume, 22(1), 102-106.

Ando, Y., Araki, S., Shimoda, O., & Kano, T. (1992).

Role of autonomic nerve functions in patients with

familial amyloidotic polyneuropathy as analyzed by

laser Doppler flowmetry, capsule hydrograph, and

cardiographic R–R interval. Muscle & nerve, 15(4),

507-512.

Mano, T. (1990). Sympathetic nerve mechanisms of

human adaptation to environment—findings obtained

by recent microneurographic studies. Environ. Med,

34, 1-35.

BIOSIGNALS2015-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

218