Electrical Impedance Plethysmography -...

Electrical Impedance Plethysmography

A Physical and Physiologic Approach to PeripheralVascular Study

By JAN NYBOER, Sc.D. , M.D.Assisted by Marian Ml. Kreider, M.D., and Leonard Hannapel, M.ID.

The quantity of blood measured by electrical impedance plethysmography is defined by its resistiveeffect in parallel to the resistance of other tissue of the segment. By substitution of this parallelresistive value, together with data relative to the resistivity of blood and the length of the seg-

ment in the formula for the volume of an electrical conductor, we are able to derive the volumeof the pulse in cubic centimeters. It follows that the volume displaced from the venous reservoirand the rate of refilling of the venous reservoir of an extremity may also be determined quantita-tively.

N UMEROUS attempts and methods havebeen evolved to measure pathologicand functional vascular change in the

extremities. Little attention, however, has beengiven to the quantitative changes recordable byvarious methods of electrical plethysmogra-phy,'-5 which until recently have been difficultto formulate. The current presentation dealsprincipally with the calculation and interpre-tation of volemic changes in the extremities asthey appear measurable by electrical conduc-tivity.The electrical conductivity method gives a

physical measure of the ionic conduction of agiven body segment in contrast with electronicconduction characteristic of metallic sub-stances. An attempt will be made to restate andformulate the laws of electrical conduction asthey appear applicable to changes within abody segment which is being studied by thepassage of a radio frequency current. As hasbeen shown elsewhere,6 transient and staticvalues of electrical conductivity are associated,

From the Department of Physiological Sciences,Dartmouth 1Medical School, Hanover, New Hamp-shire and Veterans Administration Hospital, WhiteRiver Junction, Vermont.

Published with permission of the Chief 1ledicalDirector, Department of Medicine and Surgery, Vet-erans Administration, who assumes no responsibilityfor the opinions expressed or conclusions drawn bythe author.

811

respectively, with dynamic and balanced con-ditions of arteriovenous blood volume differ-ences within a given segment.

METHODA differential electrical impedance bridge oper-

ating at 150 to 200 kilocycles, as described else-where,6 is used as the basic instrument for measure-ment of small and large variations in volume of theextremities.

In this study, our measurements on human ex-tremities are made with aluminum foil strips 1 cm.wide, mounted on a gummed tape 2 cm. wide. Afterwetting the foil with a concentrated salt paste (non-drying), the electrodes are applied firmly withoutpreliminary massage or washing. This tends to avoidlocal circulatory reactions which may modify ourmeasurements. The electrodes are usually appliedcircumferentially and held loosely by a clip lead toavoid constriction (fig. 1).

There appears to be an inherent advantage in afour electrode technic as shown later. The two outerelectrodes (I, and 12) are used for applying the cur-rent, and the two inner electrodes are used for de-lineating the segment under measurement. The in-ner edge of the El,, E2 pair is the effective edge ofthe segment for measurements of the limb. ?XJostof the measurements reported here are based on aconventional two electrode method represented bythe E,, E2 pair to which the current is also appliedunder such conditions.The standard for comparison of the segmental

pulse volume resistivity is usually .05 to 0.1 ohmresistance. One ohm is used for comparison of largechanges in segmental blood volume. A small pel-centile change of the total substitution resistancefor the segment is also adaptable as a standard.

Circulation, Volume II, December, 1950

by guest on May 19, 2018

http://circ.ahajournals.org/D

ownloaded from

http://circ.ahajournals.org/

ELECTRICAL IMPEDANCE PLETHYSMOGRAPHY

THE PHYSICAL BASIS AND ELEMENTARYCONSIDERATIONS IN ELECTRICAL

VOLUME RECORDINGThe electrical impedance pulsation represents

a changing number of ions brought to thesegment by the arterial stream at a rate ex-ceeding the venous outflow during the cycle.The over-all change in volume of a segment isthe differential effect of expansion and empty-ing of the vascular components of the entiresegment.

It may be possible to account segmentally forthe volumetric shift of blood by considering itseffect as a variable parallel electrical shunt.Equations for the effect of parallel resistanceare well known.

FIG. 1. A photograph of the forearm illustratesthe manner of applying the aluminum foil electrodes.The arrangement is tetrapolar. The inner electrodesare designated E1, E2, and the outer electrodes IL,I2. The segment of 20 cm. in length between E1 and

E2 is the effective resistance (Ro).

The total electrical conductance of the ex-

tremity segment is probably equal to the sumof the paralleled conductances of the blood andsegment proper. Each additional pulse of bloodrepresents another path through which elec-trical current will flow.The effective parallel resistive value of the

added or displaced blood may be derived bysubstitution of measured values in the ex-

pression:RN RO

Ro- RN'or RB R

in which Ro represents the original resistanceof the segment, RN, the new total resistance.Ro- RN, which is equal to AR, represents the

change in resistance incurred by the changein blood volume of the segment, by pulsation or

otherwise. When small volume and resistive

changes occur, then Ro expresses essentiallyvalue of the product RNRO.The volume of blood within the segment is a

direct and linear function of electrical con-ductivity. This is true within wide limits ofexpansion of elastic cylinders such as arteries,veins, intestines, and rubber tubes, as shownelsewhere6 by one of us (J.N.). The inclusionof ground meat, long bone, or ground bonechanges the slope of the relationship but doesnot destroy the lineal effect.The change in volume of blood uniformly

distributed within a segment may then becalculated from the derived expression of thevolume of a cylindrical conductor:

12VB = PR

RB(2)

in which p represents the specific resistivity ofthe segmental blood, 1 the length of the segmentbeing measured, and RB the calculated effectiveparallel resistance of the blood related to thechange. Compare equation 1 in Medical Physics,Vol. II, p. 738.

Direct comparison of the impedance andsensitive mechanical plethysmographic recordsof the volume pulse of the finger publishedelsewhere5' 6 shows no apparent discrepancybetween the forms of the curves recorded atapproximately the same amplitude. They ap-pear identical, and this suggests the similarityof origin to the volume pulse of the segment.

THE MEASUREMENT OF THE VOLUME PULSEThe volume of blood pulsed into a peripheral

segment is usually equal to the volume ofblood leaving during the cycle. If one had anaccurate measure of either the input or output,or of both volumes, it is probable that valuabledata covering vascular responses could be scien-tifically expressed. If a measure closely propor-tional to the absolute pulse volume were ob-tained, it would not be necessary to know eitherphase of the volume change.

Segmental blood flow can be approximated atpresent by the venous occlusion plethysmo-graph.7 8 This method is objectionable sincethe pressure of occlusion may modify the vascu-lar responses and other factors beyond reason-able control. If, under these conditions of

812



ownloaded from


JAN NYBOER

measurement, one also knows the pulse rate,the pulse volume may be deduced from suchdata.

Since there is usually a continuous venousrun-off from the segment during the cycle, it isapparent that the recorded volume pulse is thenet difference between a momentary excessivearterial input and the run-off of blood, whetherit be venous or arterial, during the cycle in theabsence of stasis.

If the above events are true, the recordedpulse volume is directly proportional, but notnecessarily equal to the sum of the true arterialinflow and the venous outflow from a givensegment. It follows that the mean height meas-urement6 of the pulse wave should be a validindex of this proportional volume. In our study,twice the mean height for the area under thecurve is chosen to represent both input andoutput volume. In effect, this represents asequestration of the total segmental input with-out occlusion or run-off of the venous return.

In practice, one obtains the mean height ofthe pulse volume by planimetric integrationover the entire pulse distance. The measure-ment of several pulses serves to reduce theerror. This value is multiplied by two, since therecorded volume increase served both as ameasure of input and of output volumes duringthe entire pulse cycle.

This measured value is then interpreted interms of ohmic resistance based on comparisonwith the 0.1 ohm standard recorded in theoriginal trace or in the records with the equiva-lent substitution of resistance. This value inohms represents AR of equation 1. As one hasmeasured the approximate substitution resist-ance (Ro) for the extremity conductor beforethe pulse volume modified the conductor, onemay then derive RN arithmetically.The effective parallel shunt, RB, is then calcu-

lated from the known values (RN, Ro, andRo- RN) by equation 1.The calculation of pulse volume is completed

by equation 2. In addition to RB, one mustknow the other measurable units defined in thisequation. We will assume that the measuredresistivity of nonflowing venous or arterialblood will supply the factor p. Theoretically,this may be different for each individual. Its

value is in the order of 145 i 3.73 S.E. ohmcm. at 37.5 C. measured at 173 kilocycle fre-quency. We assumed a value of 150 ohm cm.for many of our measurements. The derivedpulse volume multiplied by the heart rate givesthe proportional volume of expansion of thesegment in cc. per minute. If one knows thevolume of the segment, it is possible to expressproportional expansion in terms of volume perminute per 100 cc. of segment.The volume of the peripheral segment is

never an ideal cylindrical conductor for whichthe equations were designed. Some correctionsshould probably be made for the shape of theconductor. Cole and Curtis9 have discussed thisfactor in relation to cell models.

Comparative measurements of similar seg-ments on opposite limbs obviate many of thedifficulties due to shape and enhance the valueof our observations in a limited number ofcases.°0

RESULTSBipolar and Tetrapolar Resistive Measurements

of a Body Segment Including the SkinHorton and van Ravenswaayll and Barnett12

have shown that the reactive component ofelectrical impedance is very high in skin com-pared to its value in deep tissues. If this is true,then it appears that the determined values forimpedance of a segment in which the deeptissues predominate in volume are probablytoo high by bipolar measures. Since the totalimpedance of the segment enters into the calcu-lation of the parallel resistance, the value ofmean pulse volume would probably be in error.

In table 1, column Ro shows that for equalsegments, the four electrode method gives con-sistently lower values for total segmental re-sistance than the two electrode method. Noattempt was made to measure at controlledconditions of rest, exercise, postprandial time,or room temperature. The subject was re-cumbent.The four electrode method basically elimi-

nates the highly reactive skin and leaves onewith a better measure of the internal tissues,including the blood, which appear predomi-nantlv resistive to alternating current. Thepulsatile volume should probably be calculated

813



ownloaded from



on the resistive values obtained by tetrapolarleads, if a closer approximation to the trueproportional pulse volume is desired.Column Ro - RN or AR represents twice the

mean height of the pulse expressed in ohms. Itappears that no consistent prediction can bemade of this value under the conditions ofmeasurement (but at no time do these exactlyequal each other).The value, however, for the parallel resistive

effect (RB) of the blood pulse volume is pre-dictable and consistent. Each of the tetrapolar

legs electrically shunted in parallel are shown(table 2). The value for the total resistance ofboth legs is predicted to be 38.1 ohms; how-ever, by measurement, we obtained 37.2 ohms,which is reasonable. For mean change in pulseheight, we predicted .0187 ohm and obtained.0191 ohm. For calculated volume changes, wepredicted .773 cc. and obtained .828 cc. forpulse volume. We predicted 47.4 cc. and ob-tained 50.8 cc. for pulse volume per minute.Similar values are shown for predicted andmeasured parallel resistances and parallel pulse

TABLE 1.-Comparative Resistive Values of Pulse Volume by Bipolar and Tetrapolar Measurement*Four Electrode Studies

Segment Length Rs RNt Re - Rg RB X 105 Pulse Pulse Minute Room Time(Recumbent) (cm.) (ohms) (ohms) (ohms) (ohms) Volumet Rate Volume Temp. (min.)(cc.) (cc.) (C.)

Right forearm ..... 20 86.7 86.58 .120 0.626 .959 56.6 54.3 27.5 3Left forearm .. .... 20 95.9 95.789 .111 0.824 .728 56.6 41.2 27.5 9Right forearm..... 20 90.7 90.611 .0894 0.898 .668 54.1 36.1 27.3 14Left forearm.... 20 96.1 96.024 .076 1.214 .494 54.1 26.7 27.1 18

-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Two Electrode Studies

Right forearm ..... 20 99.5 99.383 .117 0.844 .711 57.7 41.0 27.5 0Left forearm ....... 20 108.5 108.38 .115 1.02 .589 56.6 33.3 27.5 5Right forearm.. ... 20 103.6 103.50 .0986 1.09 .552 56.6 31.2 27.3 11Left forearm....... 20 110.2 110.12 .0769 1.58 .380 57.7 21.9 27.0 21

* In this experiment a progressive change in measurements is shown with time. This is probably physiologicrather than physical in origin. Premeasurement period of rest may have been inadequate. The decreased minutevolume with time is suggestive of peripheral vasoconstriction which occurs with the change to a recumbentposture.

t Probably no more than three or four figures at any time are significant in columns designated RN in this andsubsequent tables. They merely illustrate the influence of the shunt due to the pulse. For practical purposes incalculation of pulse volume, RN equals Ro.

t Resistivity of the blood assumed as 150 ohm cm.

measurements is lower than the correspondingbipolar measurement of a given arm.The resultant effect on the calculation of

pulse volume shows that tetrapolar measure-ments give higher values than bipolar measure-ments.

It follows that the values for pulse volumesper minute in our data are significantly higherfor the tetrapolar method in spite of slightlyhigher pulse rate during the bipolar measure-ments in the reported experiment (table 1).

Parallel Impedance of Two Similar Segments as aMeasure of Parallel Pulse Volume

The effect of electrically parallel circuits isdefined by equation 1. The resistive values forstudy of the left leg, the right leg, and of both

volume studies in the arm segment (table 2and figure 2).A bipolar technic was employed because the

four electrode method was not available at themoment. Perhaps a better correlation may beobtained with more careful control. The experi-ment proves that we may study complex bio-logic electrical impedances, such as the bloodpulses, in parallel, and predict the final im-pedances and the parallel segmental expansiondue to blood with fair accuracy, under changingphysiologic conditions.

External Parallel Shunts on the Skin between theEffective ElectrodesThe summary of such an experiment is tabu-

lated and shown (table 3 and fig. 3). It is noted

814



ownloaded from


JAN NYBOER

TABLE 2.-Parallel Imipedance Study of Two Similar SegmentsStudy of Twvo Legs on Subject J. N.

Segment Length Ro(Recumbent) (cm.) (ohms)

Left leg .............................. 20 76.5Right leg ............................ 20 75.8Both legs in parallel-measured ...... 20 37.2Both legs in l)arallel predicted ...... 20 38.1

Ro -RN RB X 103 Pulse

(ohms) (ohms ) Volume*

~(cc.)

.0382 1.53 .392

.0367 1.57 .392

.0191 0.725 .828

.0187 0.776 .773

Study of Two Forearms on Subject M. K.

Right forearm ....................... 20Left forearm ......................... 20Both forearms in parallel-measured. 20

20Both forearms in parallel-predicted. .20

144.5 .1069157.9 .161774.7 .078174.7 .061775.5 .0650

1.951.540.7140.8310.861

.307

.390

.844

.721

.697

* Resistivity of blood assumed as 150 ohm cm.

RA

LA

RA

LA

RA

LA

sVYV

pz

C',qp % 0.1 ohm

FIG. 2. The parallel effect on the pulse volume ofcombining resistive changes in the right arm (RA),and left arm (LA), is shown in the third and fourth

curves. The calculations are outlined in the text andtable 2, subject M.K. An A.C. coupled amplifier con-

trols the terminal oscillograph. The drift shown in

the standardization curve is characteristic of the end

recording equipment.

that the total resistance (Ro) of the leg segmentdrops from 75.8 ohms to 71.7 ohms on applica-

tion of salt paste between the electrode region.A further drop to 52.9 ohms occurs on wrappinga sheet of aluminum foil around the mid-section of the measured segments. Later, bycalculation, it appeared that the effective paral-lel resistance of the salt and aluminum foil(surface shunt) was 175 ohms.The influence of salt paste on the final pulse

volume because of its coldness appears to haveresulted in a slight reduction of 1.3 cc. per

minute from a level of 24.0 cc. per minute forthe segmental expansion.The addition of aluminum foil over the salt

paste was followed by measurement six minuteslater. These values showed an increase in pulsevolume and minute volume. The level exceededthe paste study by 6.1 cc. and the control levelby 4.8 cc. The change was not due to room

temperature, which was constant. The subjectcomplained of a feeling of warmth under thealuminum foil. The change might have beendue to local changes interfering with radiationof heat from the segment. It is improbable thatthe physical application of a shunt resistanceof 175 ohms in parallel with the leg was thecause of physiologic changes. It is concludedthat physiologic changes in pulse volume andminute volume are secondary to the peculiarmethod of producing the physical shunt.By further experiment under controlled con-

ditions, an external physical shunt of 209 ohmsparallel to the blood circulation and the tissuesof an extremity did not produce a physiologicchange, but the anticipated physical effect of

PulseRate

61.361.361.361.3

MinuteVolume(cc.)

24.024.050.847.4

Room TmTemp. (mi.)

21.0 020.8 4821.0 4421.0 24

75.065.269.770.169.9

23.025.458.851.648.7

23.023.023.023.023.0

0368

1.5

8 1 5

j



ownloaded from


816 ELECTRICAL IMPEDANCE PLETHYSMOGRAPHY

TABLE 3.-The Effect of a Low-Resistive Parallel Shunt on the Skin Between the Electrodes on the Resistancesand Volume Pulse

Segment Length Ro Ro - Rg RB X 105 Pulse Pulse Minute Room Time(Recumbent) (cm.) (ohms) (ohms) (ohms) Volume Rate Volume Temp. (min.)____________________________________ ~~~~~(cc.) (cc.) (C.) (m.

Right leg (control) ....... ...... 20 75.8 .0367 1.57 .392 61.3 24.0 20.8 0Right leg covered with paste... 20 71.7 .0318 1.62 .370 61.3 22.7 21.0 8Right leg covered with alumi-num foil ..................... 20 52.9 .0223 1.25 .479 60.0 28.8 20.8 14

Right leg after aluminum foiland paste removed ...... ..... 20 77.8 .0373 1.62 .371 63.8 23.6 21.0 32

The calculated effective parallel resistance of the aluminum foil and the electrode salt paste is 175 ohms.Subject (J. N.) stated that the paste felt cold while the aluminum foil covering the leg produced a warm

feeling which may account for some shift in minute volume.

2

3

FIG. 3. The shunting effect, between bipolar elec-trodes, of smearing a salt jelly over the skin is shownby reduction of pulse amplitude (2) in comparisonwith control pulses (1) from the right lower leg. Astill greater surface shunt is produced by addingaluminum foil about the segment (3). All curves arerecorded at the same sensitivity of the instrument.See text and table 3.

reducing the recorded pulse volume was evi-dent and predictably reproducible.

Mild Exercise and Pulse Volumes of the ForearmSegment

Subject E. T. was studied before and afterone minute of flexion and extension exercise ofthe fingers, using the bipolar method on theforearm. The results are shown (table 4) andindicate a significant rise in the calculated pulsevolume and minute volume by impedancemethods on the same arm. There is an evidentreturn to normal pulse volume and minutevolume within six minutes. A change from 19.9cc. to 26.0 cc. per minute is shown after oneminute of exercise of the same type in oneforearm of subject J. N.

It appears from limited data that the vXascu-lar expansion may be differentiated as toexercise in the extremities as shown in theforearm segment under controlled conditionsby electrical impedance studies.

TABLE 4.-Physiologic Response in Pulse Volume to Exercise*Subject E. T.

Subject(Recumbent)

Forearm before exercise.......Forearm after exercise........

Forearm before exercise ........Forearm after exercise ........

Length(cm.)

20202020

2020

Ro Ro - RN RB X 1us Pulse(ohms) (ohms) (ohms) V olumet

(cc.)

132.0 .0713 2.44 .246126.0 .0874 1.82 .330128.0 .0862 1.90 .316128.5 .0736 2.24 .268

Subject J. N.

PulseRate

56.257.757.757.7

MinuteVolume(cc.)

13.819.118.215.4

91.8 .0437 1.93 .311 63.8 19.9

94.1 .0640 1.38 .433 60.0 26.0

RoomTemp.(C.)

20.020.020.320.3

Time(min.)

012.56

* Exercise was moderate flexion and extension of the fingers for one minute.t Resistivity of the blood assumed as 150 ohm em.

20.8 021.2 1



ownloaded from


JAN NYBOER

Electrical Measurement of Large Changes inVenous Blood Volume of the Forearm SegmentUntil recently, it has been difficult to approxi-

mate the volume of blood leaving or enteringan extremity segment passively, such as mayresult from posture. On the basis of the forego-ing changes in resistance produced by the pulsevolume, we reanalyzed the type of experimentoriginally recorded at Wright Field.A forearm segment was prepared as usual

with two electrodes for resistive studies at agiven posture. The arm was allowed to rest ina natural position and Ro was measured bysubstitution. Ro - RN or AR was recorded at low

the initial effect was to produce the wave in theconductivity record opposite region 5. This mayhave been associated with partial reflux orretrograde flow of venous blood which was notheld back by the venous valves above the proxi-mal electrode on the arm.

Thereafter, the return curve was more uni-form at region 6 before rising asymptoticallytoward the baseline at regions 7 and 1. Thevolume of the arm and the venous reservoir hadnearly refilled to the original level at regions 7and 1, judging from the conductivity curve.Region 6 may be important as a measure ofunobstructed filling of the venous reservoir.

TABLE 5.-The Effects of Raising the Arm on the Electrical Resistance and Calculated Venous Outflow

Subject(Recumbent)

Left forearm

Right forearm

VenousOutflow(mm.)

10.08.511.016.020.021.019.025.027.042.042.0

Standard of1 ohm(mm.)

5.54.04.56.05.05.04.56.07.55.55.5

Ro(ohms)

118118118118118118118112112112112

AR(ohms)

1.82.12.42.74.04.24.24.24.97.67.6

RN(ohms)

119.8120.1120.4120.7122.0122.2122.2116.2116.9119.6119.6

RB X 103(ohms)

7.816.755.925.333.603.433.433.022.671.761.76

VenousOutflow

(cc.)

7.78.910.211.316.717.517.519.222.534.234.2

Room temperature 27 to 27.5 C.; arm volume approximately 650 cc.; resistivity of blood assumed to be 150ohm cm.

Data arranged in the order of magnitude for AR of the given segment.The linear relation to the volume of venous outflow becomes self-evident.

recording sensitivity, so that 1 ohm was equalto 5 mm. excursion of the galvanometer. Thisis shown in figure 4 at region 1, where there isno evidence of drift or other imbalance exceptthe standard. The base of measurement was 165ohms and the pulses were visible but not meas-urable. The arm was then raised graduallyabove the shoulder level as the subject wassitting. This was associated with a precipitousdrop in electrical conductivity and presumablya passive decrease in the amount of venousblood stored in the veins of a given segmentThis emptying of the veins is recorded oppositeregion 2. This was followed by a very slow rateof decrease in conductivity opposite regions 3and 4 of the record.When the arm was lowered moderately fast,

It was estimated from the measurements thatthe drop in 7 ohms impedance was equivalentto 14.5 cc. of venous outflow from the segmentby raising the arm under these given conditions.As the displacement of the segment was about650 cc., the change in volume was calculated tobe 2.23 cc. per 100 cc. of a given arm segment.Other typical results of partially controlledmeasurement during recumbency of a given sub-ject, arranged according to degree of deviationin volume, are in table 5.Record B demonstrates how repeated raising

and lowering of the arm produces changes inconductivity similar to regions 1 to 7 on recordA. N7o attempt to measure these has been madeas more rigid controls of activity and posture

817



ownloaded from



must be defined in order to quantitate thethe events properly.

Electrical MIeasuretnent of Blood Flow lWithoatOcclusion During Rejillintg of the VenousReservoir

"Exact analysis" of physiologic change hasbeen a difficult, tedious, and ungratifying task

a series of preliminary observations (table 6).Limited body rest, inactivity, recumbency, anda special position for the arm were the controls.Room temperature and time between and duringelevations of the extremity, as well as the degreeof horizontal elevation of the proximal electrodein relation to the angle of Louis and the body,wvere determined.

TABLE 6.-The Calculated Rate of Filling of the Venous Reservoir After Rapidly Lowering the ArmWi1lithout Occlusion

Segment(Recumbent)

Left forearm

Right forearm

VenousFilling

(mm./3 sec.)

2.03.02.53.53.03.03.05.55.05.55.5

Standard of1 ohm(mm.)

45.54.56.05.05.04.57.55.56.05.5

Ro(ohms)

118118118118118118118112112112112

AR RN

(ohms) (ohms)

0.500.550.550.580.600.600.670.730.910.921.00

117.5117.45117.45117.42117.4117.4117.33111.27111.09111.08111.0

Room temperature 27 to 27.5 C.; arm volume approximately 650 cc.; resistivity of blood assumed to be 150ohm em.

The data as arranged shows the linear relationship between AR, the resistance change per unit time and thecalculated rate of return or filling of the venous reservoir with blood.

FIG. 4. The effect of raising and lowering the arm

on the recorded electrical impedance of a segment ofthe right forearm (curves A and B). The electricalconductivity decreases and the volume of the forearmdecreases as the arm is raised. See the text for furtherdetails and interpretation. A bipolar arrangement ofelectrodes was used for these procedures. A stringgalvanometer is the end recorder. The paper filmspeed is 6.25 mm. per second.

of measuring "blood flow" in various segmentsof the extremity by mechanical plethys-mography following venous occlusion at theoutflow region of a segment.7' 13Taking advantage of the technic described in

the previous section of this paper, we recorded

At region 6 of the experimental curves (fig.4), we assumed that the fairly uniform rate ofchange was an index of the rate of flow ofblood, refilling the recently collapsed venous

reservoir of the arm segment.We derived the results in table 6 by measur-

ing AR for a portion of slope 6 as representativeof the portion of blood volume (AV) which isfilling the veins in a given time (AT). It is alsonecessary to know the equivalent parallel re-

sistance for value AR, as well as the length ofthe segment and resistivity of the blood. Underthese circumstances, substitution of values inequation 2 should permit the calculation forthe change in volume (AV). In the illustration(fig. 4), it appears that the volume is 1.75 cc.

for a period of three seconds. This represents a

flow of 35 cc. per minute. This figure is reason-

able for the segment of about 650 cc. volume.Further results and calculations on the same

subject under better controlled conditions are

arranged in order of magnitude in table 6. Itappears justified to conclude that with our

RB X 10(ohms)

2.802.542.542.402.332.332.101.741.391.381.27

Net VolumeFlow

(cc./min.)

42.947.847.850.051.551.557.669.386.387.094.8

818



ownloaded from


JAN NYBOER

technic it is possible to measure the rate ofvenous filling after the venous reservoir of theextremity has been collapsed by posturalmeans. Venous occlusion may be done while theextremity is raised, in order to prevent venousoutflow. UInder such conditions, blood accumu-lates, however, immediately, and the slope oftenapproaches the slope described in region 6(fig. 4).

It would appear that the new technic is morephysiologic and potentially as valid a measureof peripheral flow as derived by venous occlusionmethods [Brodie and Russell (1905),13 Hewlettand van Zwaluwenburg (1909) ,7 and the wellknown modifications of these]. A comparativeanalysis of simultaneous impedance methods ofdetermining blood flow is being reported else-where.7', 10

DISCUSSION

The electrical impedance method under con-sideration here is not basically different fromthat employed 10 years ago.4 Largely because ofmisunderstanding of the basic principles in-volved, progress of impedance measurement ofthe multicellular organism and its circulationhas been slow and involved.

It is reasonable to assume that if there areseries and parallel resistive circuits in the organ-ism, then some of its functions may almostsurely be defined in terms of resistive changes.The peripheral circulation is one of these func-tions. If the blood pools indefinitely or tran-siently within a limited segment, its resistivitywR7ill be added in some manner to the resistivityof the segment. Continuous records of electricalconductivity of a body segment trace the courseof a blood volume pulse and measure its magni-tude by virtue of the number and mobility ofions. If one is still in doubt on this fact, heshould follow the blood conductivity of anartery in animal or man before, during, andafter injection of electrolytic or nonelectrolyticsubstances. The dilution curve is clearly re-corded elsewhere.6 The data derived from suchan observation appears to be a function ofpartial or complete circulation time, cardiacstroke volume, total cardiac output, and bloodvolume, and diffusion from the vascular system.The various functions of the peripheral circu-

lation which may be defined also by impedance

measures are pulse volume, minute pulsevolume, pulse velocity, the rate of change ofpulse volume, pulse form, blood pooling andemptying, particularly of the venous reservoirs,and secondarily, the backflow into the reservoirassociated with postural changes in the ex-tremity. The rate of filling of the venous reser-voir from the arterial side may be a function ofsegmental blood flow, as observed electrically.This long list of possibilities sounds like a copyof the encyclopedia, but in all fairness, weshould find out why and where we have failedor lost our patience to define a given vascularfunction by electrophysical measurement.One basic clue to evaluation of the volume

of blood at any given moment in a given seg-ment is its parallel resistive measure. To findthis, we must have other data as outlined.The concept that volume of substance is a

direct function of its electrical conductivity istaught in elementary physics of conductors inrelation to electronic conduction. There is nojustification in believing that it does not applyto ionic conductors, such as the blood. If theblood does not alter too rapidly in its elec-trolyte and nonelectrolyte ratio, its con-ductivity should justifiably enter into volumedetermination as it is directly proportional tothe measure of segmental volume.6 It is notdifficult to measure in vitro. If the body seg-ment is measured at a given frequency, theblood in vitro relative to the segment shouldbe measured at this frequency unless we havereached the infinite frequency impedance ofthis tissue with our signals. The physiologicconditions of flowing blood are more difficult toreproduce.The vascular reactions to drugs, nerve block,

arterial and venous occlusion, exercise, posture,anxiety, pleasure, physical agents, trauma, sur-gery, disease, and stimulation are but a few ofthe responses which may be carefully evaluatedto advantage by electrical impedance methods.Burch's8 sensitive mechanical plethysmographhas already proved to be very useful in thisregard, but Goodyerl" has recently pointedout some advantage to the electrical impedancemethod, although he did not show how to exactthe volumetric data from his records in con-ventional terms.

819



ownloaded from



Recently, Coulter and Pappenheimer'5 havefound that turbulence in flowing blood does notinfluence electrical impedance measurement ofblood. Thus blood cells remain oriented inturbulent flow. At low flows, however, theblood viscosity decreases with increasing flow.This effect parallels the observation of Velickand Gorin'6 who observed that the electricalresistance of flowing blood measured in thedirection of flow was less than that of blood atrest. At present, we cannot supply a correc-tion for the resistivity of blood, if a correctionshould be made for its flow or given velocity.Some of the other major potential disadvan-

tages concerned with conductivity methods arethe unknown mean temperatures of the segmentand its blood streams. Thermal gradients'7 arepresent in the arteriovenous vascular system.This presupposes differences in the conductivityof arterial and venous blood of the peripheralsegment. In addition to this, the relative veloci-ties'8 and proportion of plasma and red cellsvary throughout the vascular system, and,therefore, are influential in distributing changesin resistance. Fluid shifts between the vascularand extravascular compartments also modifythe electrolyte-nonelectrolyte ratio of each andincidentally the paralleled resistive effects per-taining to each. Pickering and Dow'9 ascribeconsistent findings of higher relative cellvolumes and plasma protein levels in venousblood than in simultaneous arterial samples tothe arteriovenous shift of water from plasma tocells. These factors must be evaluated ulti-mately in pulsatile and nonpulsatile impedancestudies of body segments, if this approach is tobecome a useful tool in evaluation of vascularphenomena.

SUMMARY

1. The quantitative measure of pulse volumeand the venous blood pool is defined in termsof its parallel electrical resistance to the re-sistance of the whole segment or of the deeptissues of the segment.

2. The parallel resistivity of a transientchange in blood volume is currently evaluatedby bipolar or tetrapolar methods of conductionmeasurement.

3. The parallel resistive value of a given

pulse or a shift in segmental blood volume,together with data of the linear dimensions ofthe segment and the resistive value of the bloodare entered into the equation for the volumeof an electrical conductor to calculate the quan-titative volume of displaced or new blood.

4. The pulse volume is proportional, but notnecessarily equal to the true arterial inflow orvenous outflow from a segment. It follows thatthe minute pulse volume is proportional to thesegmental blood flow.

a. The physical effect of parallel circuit ar-rangement of both legs by experiment provesunequivocally the quantitative nature of thesegmental blood pulse volume as evaluated byits parallel resistive effect.

6. It appears that the rate of blood flow in aperipheral segment may also be approximatedby evaluation of the rate of filling of a previ-ously emptied venous reservoir produced by achange in posture of the segment.

7. The electrical impedance methods appearsufficiently accurate and sensitive to warrantfurther consideration for application to basicand clinical medical science.

ADDENDUM

Since the original submission of our manu-script, F. H. Bonjer,20 in his thesis "Circulati-eonderzoek door Impedantiemeting," has in-dependently correlated the recently publishedequations' and arrived at the quantitative solu-tion for pulse volume from electrical impedancedata on extremity segments, i.e.:

Al? 12AV= .tR -

Ro Ra

in which AV7 represents the pulse volume re-lated to its maximum excursion and the othervalues as defined above. In effect, this is equiva-lent to equation 2.At present, his impedance unit employs a

frequency of 60,000 cycles and a percentilechange in resistance for a standard of compari-son. A tetrapolar electrode system is also usedto his advantage. Open ring flexible electrodeswere made from braided fine copper wire.He also refers to Russian, French, German

and Austrian investigations on body im-pedance, some of which were unknown to us.

820



ownloaded from


JAN NYBOER

Our earliest pulsatile impedance investigations'reported in 1940 were wholly independent ofsimilar contributions published elsewhere.

ACKNOWLEDGMENTThe authors are greatly indebted to the Staff of

New York Postgraduate Hospital and MedicalSchool, formerly Columbia University, for the ini-tial promotion and development of this researchand to its special cardiac committee for financial-aid; to Dr. Kenneth S. Cole for his encouragement;to Samuel Bagno for his continued skillful assist-ance and advice; to Mr. Walter Rahm for his phil-anthropic support and technical efforts; to Dr.Leslie Nims for his suggestions and collaborationwhile the project was at Yale Aeromedical Unit,New Haven; to Mr. 0. L. Bard of Detroit for hisgenerous gift and encouragement; to Margaret DayBlake, Louisa P. Burling, and Edward B. Burlingof Cornish, New Hampshire, for their generosityand vision; to Dr. Allen L. King and Dr. Willis M.Rayton, of Dartmouth College, as consultants inphysics; to Dr. Arthur Miller of Boston for hissuggestions; to Sanborn Instrument Company ofBoston for the loan of a tribeam oscillograph tofacilitate some of these studies; to the Staffs of theMary Hitchcock Memorial Hospital and the Vet-erans Administration Hospital of this area for theirmany services; and to all others who were of helpover the period of ten years. Dr. Avrom Barnett,now deceased, was an active participant in the re-search project when it was started.

REFERENCES1 ATZLER, E., AND LEHMANN, G.: tber ein neues

Verfahren zur Darstellung der Herztatigkeit(Dielektrographie). Arbeitsphysiol. 5: 636,1931-1932.

2ROSA, L.: Diagnostische Anwendung des Kurz-wellenfeldes in der Herz und Kreislaufpath-ologie (Radiokardiographie). Ztschr. f. Kreis-laufforsch. 32: 118, 1940.

3HOLzER, W., POLZER, K., AND MARKO, A.: RKG.Rheocardiography. A method of circulation'sinvestigation and diagnosis in circular motion.(English translation.) Vienna, Wilhelm Maud-rich, 1946.

4NYBOER, J., BAGNO, S., BARNETT, A., AND HAL-SEY, R. H.: Radiocardiograms: Electrical im-pedance changes of the heart in relation toelectrocardiograms and heart sounds. J. Clin.Investigation 19: 963, 1940.

-5 , X AND NIMs, L. F.: The impedanceplethysmograph, and electrical volume recorder.CAM Report No. 149, OSRD, June 10, 1943.

6 : Impedance plethysmograph. In MedicalPhysics, ed. 2. Chicago, Yearbook Publishers,1950. (Includes extensive bibliography on elec-trical impedance.)

7HEWLETT, A. W., AND VAN ZWALUWENBURG, J.G.: The rate of blood flow in the arm. Heart1: 87, 1909.

8 BURCH, G. E.: A new sensitive portable plethys-mograph. Am. Heart J. 33: 48, 1947.

9 COLE, K. S., AND CURTIS, H. J.: Electrical phys-iology: electrical resistance and impedance ofcells and tissues. In Medical Physics, ed. 1.New York, Yearbook Publishers, 1944. P. 344.

10 NYBOER, J., KREIDER, M. M., HANNAPEL, L.:Basic and quantitative electrical conductivitystudies of the peripheral body segments. Tr.Am. Therapeutic Soc. In press.

11 HORTON, J. W., AND VAN RAVENSWAAY, A. C.:Electrical impedance of the human body. J.Franklin Inst. 220: 557, 1935.

12 BARNETT, A.: The basic factors involved in pro-posed electrical methods for measuring thyroidfunction. IV. A combined study of the skinand deep tissues by the 2, 3, and 4-electrodetechniques. West. J. Surg. 45: 612, 1937.

13 BRODIE, T. G., AND RUSSELL, A. E.: On the de-termination of the rate of blood flow throughan organ. J. Physiol. 31-32: 45, 1905.

14GOODYER, A. V. N.: Observations on the im-pedance plethysmograph. Proc. Am. Soc. Clin.Investigation 27: (Part II) 536, 1948.

15 COULTER, N. A., AND PAPPENHEIMER, J. R.: De-velopment of turbulence in flowing blood. Am.J. Physiol. 159: 401, 1949.

16 VELICK, S. E., AND GORIN, H. G.: J. Gen. Physiol.23: 753, 1940.

17 HORVATH, S. M., RUBIN, A., AND FOLTZ, E. L.:Thermal gradients in the vascular system. Am.J. Physiol. 161: 316, 1950.

18 FREIS, E. D., STANTON, J. R., AND EMERSON, C.P.: Estimation of relative velocities of plasmaand red cells in the circulation of man. Am. J.Physiol. 157: 153, 1949.

19 PICKERING, R. W., AND DOW, P.: Arterio-venousdistribution of brilliant vital red and T-1824injected into dogs. Am. J. Physiol. 161: 221,1950.

20 BONJER, F. H.: Circulatieonderzoek door Impe-dantiemeting. Groningen, Drukkerij I. Oppen-heim N. V. June, 1950.

821



ownloaded from


JAN NYBOER, Marian M. Kreider and Leonard HannapelPeripheral Vascular Study

Electrical Impedance Plethysmography: A Physical and Physiologic Approach to

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1950 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.2.6.811

1950;2:811-821Circulation.

http://circ.ahajournals.org/content/2/6/811the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document.

Permissions and Rights Question and Answer Further information about this process is available in therequested is located, click Request Permissions in the middle column of the Web page under Services.the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:



ownloaded from

http://circ.ahajournals.org/content/2/6/811

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


Electrical Impedance Plethysmography -...

Documents

Transcript of Electrical Impedance Plethysmography -...