Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Vol.

20, No 6,1998

ELECTIUCAL PARAMETERS OF HUMAN BLOOD

Yekta Ulgen and Man2 Sezdi

Institute of Biomedical Engineering, Bogaziqi University, TURKEY

E-mail: ulgeny@boun.edu.tr
sucumana@boun .edu.tr

Abstract-Blood samples taken from healthy donors are ranging fkom 31 to 50: 2.6 ml blood, 2.4 mi blood+0.2 ml
diluted with 0.9 % saline, to obtain samples with saline, and 2.2 ml blood+0.4 ml saline.
hematocrit values ranging from 31 to 50. Bio- Hematocrit is defined as the ratio of the total volume of red
impedance spectroscopy is then performed on these blood cells to the volume of the blood sample. Preparation
samples. Complex impedance measurements in the of the samples and measurements were carried out right
frequency range from 100 kHz to 1 MHz were fitted after the blood samples were taken. Whereas original
into Cole-Cole diagrams, using a LMS algorithm. Data hematocrit values were measured according to standard
were interpreted in term of an equivalent resistor- procedures, other blood+saliie solutions’ hematocrit values
capacitor circuit. In this hematocrit range, all four were calculated by making use of Equation (l), given
Cole-Cole parameters R, Ri, F, and a, are shown to be below:
a linear function of hematocrit. With increasing
concentration of red blood cells, the intracellular
conductivity is also increased while extracellular
conductivity is reduced and, Cole-Cole plots are shifted
upwards to the right, in the complex plane, indicating a
H’=H(I-Ci
( (
I-we I--
31
higher 01 and a lower F, We suggest that., parameters with Ci the volume concentration, We the osmotically
active water per erythrocyte volume, Sb the osmolality of
R, Rb F,, and a, are a potential index for clinically
evaluating blood samples. blood and S% the osmolality of the blood-saline mixture.
From reference [3] typical values are: Sb= 300 mOsmol
Keywords: Red blood cell, cell complex conductivity, /kg-H20; S’b = 285 mOsmol /kg-H20 and We=0.57 g /ml.
Cole-Cole plot, hematocrit. The impedance measurement of blood samples were
performed using the four-probe technique, to eliminate
electrode interface impedance.
I. INTRODUCTION A rectangular conductivity cell of 2cm x 0.65cm x 2cm
constructed from plexiglass was used in all impedance
A cell basically may be considered to be an intracellular measurements (Fig. 1). Two planar parallel gold-plated
fluid surrounded by a cell membrane immersed in an electrodes, each measuring 0.65 x 3 cm and separated by 2
extracellular fluid. Since the intracellular and extracellular cm were used to inject the current across the full cross
fluid have different concentrations of electrolytes, the sectional area of the sample. The stainless steel point
membrane behaves like an electrical capacitance[l.2]. The electrodes, 0.7” in diameter, and spaced 0.5cm apart,
electrical impedance of blood can be closely simulated by a were placed at the center of one of the side walls of the
circuit consisting of plasma resistance in parallel with cell cell, as voltage-sensing electrodes.
interior resistance and cell membrane capacitance. The
impedance of tissue decreases with fkequency because of +I -I
the short circuiting effect of the cellular membrane
capacitance.
In the following, electrical parameters of human blood that
are measured using the impedance spectroscopy technique,
are studied by altering the hematocrit values over the range
31 to 50.
Figure 1. Conductance Cell

II. METHOD AND MEASUREMENT

, was determined by placing a
The cell constant, k = (oZ)’’
Blood volumes of approximately 10 ml were drawn from 9 reference solution, 0.9 % saline, in the conductivitycell. To
healthy donors by hypodermic syringe, with 2 units of correct for its stray capacitance, complex admittance of the
heparin added to each sample. For the measurements, the saline was measured, since saline is purely resistive, the
samples were diluted with 0.9 % saline in different imaginary part directly gives this capacitance.’
volumes to obtain samples with different hematocrit values The electrodes were connected to a precision LCR meter,

0-7803-5164-9/98/$10.00 0 1998 IEEE 2983

Hewlett Packard 4284A. by means of a unity gain frequency a.The slope yields the quantity (1-a)The
instrumentation amplifier, that uses high-speed video op- phasors u and v are defined in Figure 3:
amps, offering a CMRR better than 50 dB, over the
frequency range from 1 kHz to 1 M H i , to provide true
four-electrode measurements. IuI = J(Wzi }r (ReIzi 1 r
+ - R, (5)

In bio-impedance spectroscopy, when complex impedance

The effective-capacitanceCeBis then determined from:
z* ( Z* = Z*.A / 1 ) of biological tissues is shown on a Cole-
Cole plot, a depressed circle is obtained with the center
located below the real axis at % and R, axis. The
depression angle @ is equal to ( a d 2 ) radians Fig. 3). The
Cole-Cole equation can be written as:

R e R-
Z* = R-+
Ric + Re c
I+ (jd"
K -Im[Z*]
Z
In this equation, a is an experimental parameter which has W
a value between 0 and 1 and shows the deviation from the
pure capacitance (where a=l). Since blood cells will have
slightly different shapes and structures, they will also have
different time constants. Capacitive effects of cell
membranes are usually lumped in a constant phase angle
impedance defined by

(3) Figure 3. The depressed Cole-Cole plot

Where K is a constant with dimensions of (Qsec-*). The

equivalent circuit is shown in Figure 2. The dc resistance
IV.RESULTS
value is equal to %, and R, is given by the parallel
combination of %, and Ric

npA
The specific (complex) impedance of human erythrocytes
suspended in plasma was measured for original blood
samples and samples that are prepared by adding 0.9 %
saline, in the frequency range from 100 kHz to 1MJ3,
since the characteristic frequency F, for blood is between
700 kHz and 1 MHz. Above this frequency, the cell
membranes become conductive and the resistance reflects
both the intracellular and extracellular fluid conductivies.
Blood samples were prepared according to: 2.6 ml blood
(original), 2.4 ml blood+0.2 ml saline, 2.2 ml blood+0.4
Figure 2. Electrical equivalent circuit of biological tissues. ml saline to obtain a wide range of hematocrit values.
The effect of changing the hematocrit value on Cole-Cole
In the Cole-Cole plot, the imaginary part is maximum at parameters, &om blood donor #1, illustrated in Figure 4.
the characteristic frequency o, which is determined from: The rms fitting error is always less than 0.5 %. As the
hematocrit increases; the exterior cell resistance R,
increases while the interior cell resistance Ri decreases; F,
(4) is reduced and a is increased.
Table 1 shows the Cole-Cole parameters, for original and
prepared blood samples. In this table, 0 and D letters refer
to original and diluted samples respectively. The donor
by plotting h
[I!)- against In o.The intersection of this

straight line with the In o axis gives the characteristic

numbers and diluted mixtures are represented by numbers;
i.e. D62 represents the blood sample from the 6" donor
prepared as 2.2 ml blood +0.4 ml saline.Results are
illustrated in Figure 5.

2984
 d I donor --e.I .donor
~

-u-?.dooor
+3dona
*4dom
+S.doOOr

-
+adoam
+7don01
&dollor

20 30 40 50
Hematocrit
Hematocrit

I.donor
0,24
1400 +Zdonor
-3doUOr
T i 1200
I *4d000r
Y,
0 1000 *SdoOOr
0,16
800 --t6don01
+7.dOna
2 0.14
I
-8.donor
0,12
20 30 40 50 60 -cllc9,doo01
Ot1 20 30 40 50 60-9don~

Hematocrit
Hematocrit
Figure 5. Model parameters& ,Ri ,Fe,and CL versus
It is clear that, hematocrit dependence of Cole-Cole hematocrit for nine data sets.
parameters is in general linear although the slopes are
slightly different for different donors. For a generalized
evaluation of the results, %, Ri, F,, and a, values are
Re(ohm)
calculated by taking the averages over the samples. Results
are illustrated in Figure 6.
The relationships between Cole-Cole parameters and H are 120
observed perfectly linear with regression coefficients r 110
between 0.98 and 1:
100

%= 3.1561H-37.7274 ; RzO.99 90
R,= 2.018H+178.6834 ; R=0.97
F,= 26.4371H+2100.147 ; R=0.98 80

a = 0.0044H+0.004; R=1.00 70
C e 4.1066H+432.2141
~ ; R=0.92
60

-H=43 Rl(ohm)

__I H=39

-H=35

50 60 70 80 90 100 110
90 -
Real (ohm)

30 40 SO 60

Figure 4. Cole-Cole pl& for differentH values of the blood sample #I. H (8)
2985
 Fc(kHz)
1400

1
le IRe(ohm)lRi (ohm)l I% @Hz) 1-cc ICeff(pF)
SO I 08 I I
I 122.2 82.1 824.12 I 0.78 I 213.35 I I
41 02 111.9 83.9 746.12 0.19 268.99
03 96.5 13.4 907.18 0.79 264.11
45 06 99.2 14.3 843.61 0.78 256.78
88.1 864.5
- 347.69
0.84
43 I 01 I 103.8 82.1 714.58
-
0.78
88.5 813.3
-
0.81 308.66
102.5 896.71 0.87
96.1 91.8 896.17
- 254.23
0.80

0.20
09
D31
98.7
86.0
70.1
77.3
1076.97
1090.20
- 0.81
0.8 1 211.05
I 1 - 1 t
40 I D61 I 83.6 80.3 916.71
95.0 1157.24

o.i61
92.0 953.21
101.5 964.92
81.6 102.5 953.21
0.14 74.5 110.8 918.93
107.8 1051.0
. y i 4322141 - 4.1oMlX R - 0.92
90.2 1130.86

I 36
D32
D6,
19.9
74.0
83.4
87.7
104.7
1306.13
1294.26
964.92
0.83
0.85
0.85
269.44
304.61
I 359.78
61.4
- 108.1
122.1
1038.24
965.1
126.0 1425.22
112.6 1344.75 240.13
I 31 I D42 I 63.2 I 118.2 I 1300.00 I 0.86 I 293.81 I
Table 1.

V. DISCUSSION REFERENCES

firom imp@= y human blood, it is

s p e c ~ o s ~ pon [l] B. L. Rigaud, L.Hamzaoui, N. Chauvw, E. Martinez, J. P. Morucci,
“Tissue Characterization and Modeling by Electrical Bioimpedance
possible to obtain information about the cell membrane and swomeay”, Trans. BME, 866-861, 1994.
the dynamics of ions and water molecules inside the cells, 121 T. x. Zhao. “Electrical lmoedance of Human Blood”. Proc of Ix.
in o& words; exm=llular and inm=llulat impedances. intConf.Electrical Bio-impedaice, 212-275, 1995.
is shown hat Cole-Cole of blood complex [3] E. D. Trautman, R. S. Newbower, “A Practical Analysis of the
Electrical Conductivity of Blood”, IEEE Trans. BME, Voi. 30,141-153,
conductivity can be used in estimating the total volume of 1983.
red blood cells in the samp1eJ.e. the hematocrit value. 141 J. G. Webster. “Electrical Impedance Tomography”, Adam Hilger,
As the hematocrit is increased from 31 to 50. the BristolandNewYork.,1990.
inuacellular decreased by 30% and the [51 I. J. Ackmann, M. A. Seitz, C. A. Dawson and L. L. Hause, “Specific
Impedance of Canine Blood“, Annals of BME, Vol. 24,5846,1996.
extracellular resistance as almost doubled. [6] A. Khig, B. Mahner. W. Tschiltschke and K. D.Rosenbaum,
A larger number of Cells implies a wider range Of cell time “Dielectric Properties of Red Blood Cells at Radio Frequencies: Influence
cons&ts indicated by a greater a value; i.e. the Cole-Cole of Medium Composition and Cell Volume Fractiol;”, Roc. of IX.
1nt.Conf.Electrical Bio-impedance, 83-86, 1995.
diagram is further depressed below the real axis. [I]T. X. Zhao, “New Applications of Electrical Impedance of Human
The effectivecapacitance Of decreases with Blood”, Journal of Medical Engineering & Technology, Vol. 20, Num 3,
increasing hematocrit since a larger number of cell ii5-120~~6.
capacitances will be in series. Cole-Cole parameters be [81 J. I. Ackmann, “Complex Bioelectric Impedance Measurement System
a index for evaluating b i d in clinical for the Frequency Range from 5% to IMHz”, Annals of BME, Vol. 21,
135-146.1993..
application.
2986