Carotid Artery Velocity Patternsin Normal and Stenotic Vessels

67

W. M. BLACKSHEAR, JR., M.D., D. J. PHILLIPS, P H . D . , P. M. CHIKOS, M.D.,

J. D. HARLEY, M.D., B. L. THIELE, M.B.B.S., AND D. E. STRANDNESS, JR. , M.D.

SUMMARY Duplex scanning prorides real time B-raode images of the carotid bifurcation Teasels alongwith a single gate pulsed Doppler flow Telocity detector. By using the B-mode output of the duplex system tomeasure the Doppler angle and spectrum analysis to measure the frequency content of the Doppler signal, in-stantaneous flow Telocity can be calculated.

Mean Telocity at peak systole was calculated retrospecthely in 68 common (CCA) and internal (ICA)carotid arteries of 39 patients who had undergone prior angiography and prospecthely in 30 arteries of 15healthy young controls. The ratio of mean peak ICA Telocity to mean peak CCA Telocity at systole(VlCA/VCCA) was below 0.8 in all 36 normal arteries and above 1.5 in all 21 high-grade stenoses of 60% orgreater diameter reduction. Sixty-one percent of 41 Tessels with less than 10 to 55% diameter redaction had avelocity ratio between 0.8 and 1.5. Only 10% of all ICA's with any stenotic lesion were incorrectly classified asnormal. VlCA/VCCA appears to be an accurate Indicator of the degree of ICA stenosis.

Stroke, Vol 11, No 1, 1980

THE MOST COMMONLY USED techniques fornoninvasive evaluation of the extracranial carotid cir-culation will detect only those lesions which reducedistal pressure and flow. These include theoculoplethysmographic1'2 and supraorbital Doppler*"*methods although the reported accuracy of these testshas varied considerably.1' *• 4~" The recent develop-ment of ultrasonic imaging techniques (table 1) hasspurred new interest in direct visualization of thecarotid bifurcation.

We use in our laboratory an ultrasonic duplexscanner which provides simultaneous real time B-mode arterial images and a single gate pulsed Dopplerflow to detect velocity changes.18 The velocity infor-mation is evaluated using a spectrum analyzer whichpermits quantification of the frequency content of thereflected Doppler signal.18-1T Although our work inthis area is ongoing, it became apparent during thesestudies that instantaneous velocity information couldbe of value in estimating the degTee of stenosis. Sincethe velocity of flow through a stenosis is related to thedegree of narrowing, this variable was investigated ina series of patients who had undergone arteriography.

Materials and Methods

I. Theoretical Considerations

As an atherosclerotic plaque progressively narrowsthe lumen of a vessel, the velocity of flow at that sitemust increase if constant flow is to be maintained (fig.I).18 Even when volume flow is reduced distal to a

From the Departments of Surgery and Bioengineering, Univer-sity of Washington, Seattle, WA 98195, and the Departments ofSurgery and Radiology, Seattle Veterans Administration MedicalCenter, 4435 Beacon Ave. S., Seattle, WA 98108.

This work was supported by National Institutes of Health GrantsNo. HL07293 and HL20898.

This material was presented in part at the 4th International JointMeeting on Stroke and Cerebral Circulation, Feb. 8-10, 1979,Phoenix, AZ.

Reprints: Dr. Strandness, Dept. Surgery, RF-25, University ofWashington, Seattle, WA 98195.

high-grade stenosis, the velocity of flow within thestenosis itself is elevated. Therefore, flow velocity isproportional to the cross-sectional area of the arteriallumen at the site of measurement. However, instan-taneous flow velocity is influenced, among otherthings, by vessel size and elasticity, myocardial con-tractility, and cardiac output, and these parametersmay vary considerably from one patient to another.Considerable overlap in absolute velocity measure-ments might thus be expected.

For the case of steady flow in a cylindrical tubewhich bifurcates into 2 branches of equal area, meanflow velocity in a branch bears a constant relationshipto the velocity in the parent tube.1' This relationshipsuggested that there may be a relatively constantrelationship in instantaneous flow velocity between aninternal carotid artery (ICA) and its parent commoncarotid artery (CCA) even though the situation at thecarotid bifurcation is much more complex. For exam-ple, the external and internal carotids are not of equalarea, they perfuse beds of different resistances, andflow is pulsatile, not steady. The ratio of ICA velocityto its parent CCA velocity could then decrease the in-fluence of the aforementioned factors affecting ab-solute velocity, since they should be relatively constantfor both CCA and ICA.

II. Study Protocol

The duplex scanner combines B-mode imaging witha 5 MHz single-gated pulsed Doppler.16 Carotid ex-amination is performed by placing the scan head onthe neck along the axis of the carotid vessels. Realtime B-mode arterial images (fig. 2) are generated on atelevision screen and are used as a guide for placementof the pulsed Doppler sample volume within thearterial lumen to assess flow characteristics. The B-mode image and the Doppler signal from each loca-tion examined are recorded on video tape. Spectrumanalysis can then be used to measure the frequencyshift of the reflected Doppler signal at any point in thepulse cycle (Honeywell fast Fourier transform spec-

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from

STROKE VOL 11, No 1, JANUARY-FEBRUARY 1980

TABLE 1 Ultrasonic Evaluation Techniques

Ultrasonic imaging Signal processing

1. Doppler flow imaginga. Continuous waveb. Pulsed

2. B-mode imaging

3. Duplex scanning

1. Spectrum analysisa. Doppler signalb. Bruit

trum analyzer and Model 1219 Visacorder). The spec-trum from a normal internal carotid is shown in figure3. High-grade stenoses exhibit higher peak frequenciesassociated with opacification of the area under thesystolic peak caused by elevated flow velocity and dis-turbed flow.17 In every case, the sample gate in theICA was positioned at the point of the maximumvelocity change.

The frequency of the Doppler shifted signal (Fd) isrelated to the velocity (V) of the red blood cells withinthe pulsed Doppler sample volume by the Dopplerequation:

CFd = 2VFo cos 0where C = constant related to the speed of sound intissue (15.4 X 106 cm/sec), Fo = frequency of the in-cident Doppler beam (5 X 106 Hertz), and0 = Doppler angle between the incident beam and theaxis of flow. Since 6 can be measured from the B-modeimage of the duplex system and Fd can be measuredfrom the output of the spectrum analyzer, the instan-taneous velocity of flow, V, at any time during the car-diac cycle, can be calculated with the data available.

Sixty-eight vessels in 39 patients who had beenpreviously studied with the duplex scanner and hadundergone subsequent selective 4-vessel multiview (bi-plane) carotid arteriography were selected for theretrospective review (Group I). Ten of these patientshad a unilateral ICA occlusion. All arteriograms hadbeen independently interpreted by 2 radiologists whowere unaware of the results of duplex scanning.Vessels were graded by percentage reduction of theirprojected normal angiographic diameter. Vessels with

QfAjV,

tQ2=

t

A2V2

J D 2

Q2=Qi

FIGURE 1. Relationship between the cross-sectional areaof an artery and its effect on volume flow and velocity whena stenosis is present. Q = volume flow, A = cross-sectionalarea, D = diameter.

FIGURE 2. B-mode image of a common carotid artery ob-tained with the duplex scanner. The dotted line marks thelong axis of the vessel. The dense white line represents theline of the incident Doppler beam with the position of thepulsed Doppler sample gate indicated by the white dot onthis line. The Doppler angle, 8, can be measured from thisimage.

wall irregularity without a definite constricting plaquewere classified as having less than 10 percent stenosis.Although a 50% diameter stenosis, which is equivalentto a 75% area reduction, is considered to be the pointat which carotid flow is reduced,20'21 in this study onlythose lesions of 60% or greater radiographic diameterreduction were grouped as high-grade, or flow reduc-ing, stenoses. Vessels from less than 10 to 55%diameter reduction were considered low-grade lesions.The reason for the distinction at this point was that forseveral of the vessels in the range of 45-55% stenosis,there was disagreement between our two radiologistsabout the exact degree of diameter reduction,necessitating arbitration. This uncertainty around thearea of 50% stenosis was reflected in the velocity data,which shows some overlap between velocity ratioscharacteristic of high-grade and low-grade stenoses.Thus, only those vessels which harbored lesions thatwere unequivocally flow reducing were grouped ashigh-grade stenoses.

Since the number of completely normal vessels inGroup I was low, a second group was studied prospec-tively, consisting of 30 arteries in 15 healthy youngcontrol subjects under the age of 35 without evidenceof arterial disease (Group II). All of these subjectswere classified as normal although none was studiedby angiography.

Doppler angles were measured from the B-modeimage of the video tapes for several recordings fromeach CCA and ICA (fig. 3). The average instan-taneous frequency at peak systole was then measuredfrom the output of the spectrum analyzer (fig. 4). Peaksystole was selected since it was a clearly definablepoint in each pulse cycle and it also was the time whenvelocity changes were most apparent.

The highest systolic velocity was calculated for each

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from

CA VELOCITY PATTERNS/Blackshear et al.

4 -

KH,

3-

2-

ni.

6 -

5 -

4 -

3 -

2 -

0

2 3Time (sec)

FIGURE 3. Spectrum from a normal internal carotid at the point marked by the arrow. Thepeak systolic frequency is low and the area under the systolic peak is clear, indicating noevidence of disturbed flow at this site.

CCA and ICA from the Doppler equation. The' ratioof the highest mean ICA velocity to the highest meanCCA velocity, VICA/VCCA, was also calculated foreach side. This ratio was plotted against the percent-age of angiographic diameter stenosis.

Results

The range of absolute velocity measurements in theICA and the CCA for each of the 3 angiographiccategories is shown in table 2. While the CCA velocitytends to decrease and the ICA velocity to increase asthe degree of stenosis becomes more marked, there isconsiderable overlap among the categories, confirm-ing the difficulty in using absolute velocity data forquantitation.

Figure 5 illustrates VICA/VCCA in the normalarteries in Group II. In none of these vessels did thisratio exceed 0.8 and in most it ranged from 0.4 to 0.7.

The relationship between the velocity ratio and thepercentage stenosis js shown in figure 6. In the 6 nor-mal vessels VICA/VCCA was again below 0.8. Com-bining these vessels with those in Group II gives atotal of 36 normal vessels in which the velocity ratio

I I I

\ 2 3TIMF (SFC)

FIGURE 4. Calculation of the average instantaneous fre-quency at peak systole, MF. Instrument effects, as well asflow disturbances, produce broadening of the pulsedDoppler frequency spectrum so that the peak frequency doesnotjeflecl the actual velocity of flow within the sample gate.MF is probably a more accurate representation of the flowvelocity at peak systole.

TABLE 2 Absolute Velocity Ranges (Cm/Sec)

Carotid lesion

Normal

< 10-55% Stenosis

60-99% Stenosis

Common carotid

28.6-178.4

33.2-121.5

47.0-114.1

Internal carotid

20.1-112.0

24.1-209.8

93.5-304.6

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from

70 STROKE VOL 11, No 1, JANUARY-FEBRUARY 1980

0.9-1

PEAKSYSTOLIC

Vic*

0.8

0.7-

0.6-

0.5-

Vcc. 0.4 -

0.3-

0.2-

0.1-

i

NORMAL CONTROLCAROTID ARTERIES

FIGURE 5. VICA/VCCA in the 30 healthy young controlvessels in Group II. V = mean peak velocity.

was shown to be under 0.8. In all 21 high-gradestenoses this ratio was above 1.5 and in most it wasmuch higher.

There were 41 vessels in the low-grade stenosiscategory and in 25 of these (61%) the velocity ratiowas between 0.8 and 1.5. In 7 (17%) of these vesselsVICA/VCCA was over 1.5 and in nine (22%) it wasbelow 0.8. However, 3 of the latter vessels were in theless than 10% stenosis category which had no definitestenotic plaque, only wall irregularity, and thus mightbe expected to have a velocity ratio much like that ofnormal vessels. Eliminating the 4 arteries with lessthan 10% stenosis from consideration improves the ac-curacy of correct classification of low-grade lesions to64.9% (28 of 37). Six of these 37 vessels (16.2%) had avelocity ratio below 0.8, placing them within the nor-mal range. In 7 vessels (18.9%) the velocity ratio in-dicated more severe disease than was seen on theangiogram. Only 6 of the 58 ICA's (10.3%) with adefinite stenotic lesion between 10% and 90% diameterreduction were incorrectly classified as normal.

Discussion

This retrospective study demonstrates the feasibilityof using velocity data obtained transcutaneously by ul-

PEAKSYSTOLIC

Vic.Vcc«

4 5 :

35-3 J O -

25-201.5

OB05-

0-

i |

• •

* i

' ' r~r-

0 <IO 10 20 JO 40 50 60 70 80 90 100PERCENTAGE ANGIOGRAPHIC

OUMETER STENOSIS ICA ( N - 6 8 )

FIGURE 6. VICA/VCCA in the 68 vessels in Group Iplotted against the percentage stenosis. V = mean peakvelocity.

trasonic techniques to estimate internal carotidstenoses. The velocity ratio, ViCA/VCCA, dis-criminated between normal, low-grade, and high-grade stenoses much better than the absolute velocitymeasurements. All of the high-grade stenoses (> 60%diameter reduction) were correctly classified, as wereall normal vessels, both angiogrammed and control.Sixty-one percent of the low-grade lesions (< 10-55%stenosis) were correctly classified, and a stenosis wasmissed in only 10% of the vessels with a definiteplaque. These results were reproducible upon repeatstudy of several of these patients.

From figure 6, VlCA/VCCA appears to increaselinearly as the degree of stenosis increases, althoughthere is some scatter in the data points. There are 4factors which may contribute to this variability. First,the ICA Doppler signal must be recorded preciselyfrom the point of maximum stenosis to detect themaximum velocity increase. Second, the configurationof the plaque may well influence the flow pattern in thestenosis. In this study no attempt was made to dis-tinguish between sharp, discrete stenoses and long,irregular plaques. Third, the degree of external carotid(ECA) stenosis may well influence the CCA peakvelocity. It is somewhat more difficult to produce areliable image from the external than from the inter-nal carotid with the duplex system, so accuratevelocity data from the ECA were not available onmany sides. The ECA should probably be taken intoaccount in future trials of this method, however.

Finally, as indicated earlier, when attempting toquantitate the degree of ICA stenosis, the error in-herent in interpreting angiograms is a problem thatmust be recognized. The inter- and intraobservervariability may approach 20% of vessels whenevaluating the ICA." This problem was most evidentin our study in several vessels with stenoses in therange of 50% diameter reduction in which our 2radiologists disagreed about the exact percentage ofstenosis. In these vessels, the anatomic data providedby the angiogram did not correlate well with thephysiologic data provided by the velocity ratio.

The simultaneous B-mode image and Dopplersignal of the duplex system is quite convenient forvelocity calculations; however, the same sort of datacould be obtained with other Doppler systems which

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from

CA VELOCITY PATTERNS/Blackshear et al. 71

include a method for measuring the Doppler angles.We have found Doppler frequency measurementsalone to often be misleading for this sort of calculationsince the proximal portion of the ICA usually coursesat a different angle than the CCA and this angle varieswith patients. The Doppler shifted frequency werecorded from the proximal ICA was usually higherthan that from its parent CCA in normal subjects;however, this apparent increase in velocity in the ICAwas an artifact induced by the smaller ICA Dopplerangle. When the correction for angle was applied, theactual decrease in ICA velocity became apparent. Theability to measure the Doppler angle for velocitycalculations, rather than relying on frequencymeasurements, is therefore important.

In making the determinations, it is critically impor-tant that the sample volume be placed in the area ofmaximum velocity change in the internal carotidartery. This was done to insure that the data weretaken from the point of greatest narrowing. Clearly, ifthis were not done, i.e. a sample was taken at somepoint distal to the stenosis, the peak velocity would belower as would the ratio used in our calculations.

Determination of the place of this sort ofphysiologic data in the evaluation of ICA diseaseawaits prospective studies, although the results of thisretrospective^ survey are encouraging. They suggestthat VICA/VCCA is an accurate and reproducible in-dicator of the degree of ICA stenosis and that thisvelocity data is more useful than Doppler frequencymeasurements. Not only can flow reducing stenoses beaccurately detected by this method, but many plaqueswhich are not large enough to reduce distal pressureand flow but which may be the site of intimal ulcera-tion can also be identified. It must be emphasized,however, that data are not yet available to show thatulceration can be detected by this approach.

In summary, duplex scanning and pulsed Dopplerspectrum analysis were used for velocity calculationsretrospectively in a group of patients studied byangiography and prospectively in a control group. Theratio of peak systolic velocity in the internal carotidartery to that in its parent common carotid artery wasfound to be an accurate indicator of the degree of in-ternal carotid stenosis, separating normal vessels fromstenotic ones and permitting accurate categorizationof most stenoses as high-grade (flow reducing) or low-grade (non-flow reducing) lesions.

References

1. Gee W, Mehigan JT, Wylie EJ: Measurement of collateralcerebral hemispheric blood pressure by ocular

pneumoplethysmography. Am J Surg 130: 121-127, 19752. Kartchner MM, McRae LP, Morrison FD: Noninvasive detec-

tion and evaluation of carotid occlusive disease. Arch Surg 106:528-535, 1973

3. Brokenbrough EC: Screening for the prevention of stroke: Useof a doppler flowmeter. Parks Electronics, Seattle, WA, 1970

4. Barnes RW: Doppler ultrasonic arteriography and flow velocityanalysis in carotid artery disease. In Bernstein EF (ed) Nonin-vasive Diagnostic Techniques in Peripheral Vascular Disease,St. Louis, CV Mosby Co 221-230, 1978

5. Barnes RW, Russell HE, Bone GE, Slaymaker EE: Dopplercerebrovascular examination: Improved results withrefinements in technique. Stroke 8: 468-471, 1977

6. Blackshear WM, Strandness DE: Angiographic imaging by ul-trasound. In Nicolaides A, Yao JST (eds) The Investigation ofVascular Disorders, Edinburgh, Churchill Livingstone (inpress)

7. Blackshear VM, Thiele BL, Harley JD et al: A prospectiveevaluation of oculoplethysmography and carotid phono-angiography. Surg Gynecol Obstet 148: 201-205, 1979

8. Gross WS, Verta MJ, Von Bellen B et al: Comparison of nonin-vasive diagnostic techniques in carotid artery occlusive disease.Surgery 82: 271-278, 1977

9. McDonald PT, Rich NM, Collins GJ et al: Dopplercerebrovascular examination, oculoplethysmography, andocular pneumoplethysmography. Use in detection of carotidocclusive disease: A prospective clinical study. Arch Surg 113:1341-1349, 1978

10. McDonald PT, Rich NM, Collins GJ et al: Ocularpneumoplethysmography: Detection of carotid occlusive dis-ease. Ann Surg 189: 44-48, 1979

11. Satiani B, Cooperman M, Clark M, Evans WE: An assessmentof carotid phonoangiography and oculoplethysmography in thedetection of carotid artery stenosis. Am J Surg 136: 618-621,1978

12. Moore WS, Bean B, Burton R, Goldstone J: The use ofophthalmosonometry in the diagnosis of carotid artery stenosis.Surgery 82: 107-115, 1977

13. Lye CR, Sumner DS, Strandness DE: The accuracy of thesupraorbital Doppler examination in the diagnosis ofhemodynamically significant carotid occlusive disease. Surgery79: 42-45, 1976

14. Towne JB, Salles-Cunha S, Bernhard VM: Periorbital ul-trasound findings. Arch Surg 114: 158-160, 1979

15. Barber FE, Baker DW, Strandness DE et al: Duplex scanner II:For simultaneous imaging of artery tissues and flow. UltrasonicSymposium Proceedings, IEEE Cat #74 CHO 896-ISU: 1-5,1974

16. Baskett JJ, Beasley MG, Murphy GJ et al: Screening forcarotid junction disease by spectral analysis of Doppler signals.Cardiovasc Res XI: 147-155, 1977

17. Blackshear WM, Phillips DJ, Thiele BL et al: Detection ofcarotid occlusive disease by ultrasonic imaging and pulsedDoppler spectrum analysis. Surgery (in press)

18. Strandness DE, Sumner DS: Hemodynamics for Surgeons.New York, Grune and Stratton 96-119, 1975

19. McDonald DA: Blood Flow in Arteries. Baltimore, Williamsand Wilkins 17-54, 1960

20. May AG, Van de Berg L, DeWeese JA, Rob CG: Criticalarterial stenosis. Surgery 54: 250-259, 1963

21. Berguer R, Hwang NHC: Critical arterial stenosis: Atheoretical and experimental solution. Ann Surg 180: 39-50,1974

22. Chikos PM, Harley JD: Unpublished data, 1978

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from

W M Blackshear, D J Phillips, P M Chikos, J D Harley, B L Thiele and D E Strandness, JrCarotid artery velocity patterns in normal and stenotic vessels.

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1980 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.11.1.67

1980;11:67-71Stroke. 

http://stroke.ahajournals.org/content/11/1/67World Wide Web at:

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

  http://stroke.ahajournals.org//subscriptions/

is online at: Stroke 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 available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Stroke Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on May 7, 2018

http://stroke.ahajournals.org/D

ownloaded from