Effect of Doppler Angle in Diagnosis of Internal Carotid Artery Stenosis

Effect of Doppler Angle in Diagnosis
of Internal Carotid Artery Stenosis
Abbreviations
AUC, area under the curve; CCA, common carotid
artery; CDU, color duplex ultrasonography; CI, confidence
interval; DSA, digital subtraction angiography;
EDV, end-diastolic velocity; ICA, internal carotid artery;
ISB, intrinsic spectral broadening; PSV, peak systolic
velocity; PSVr, ICA-to-CCA PSV ratio; SRU, Society of
Radiologists in Ultrasound
Received March 22, 2006, from the Department of
Radiology, Türkiye Yüksek Ihtisas Hospital, Sihhiye-
Ankara, Turkey. Revision requested April 20, 2006.
Revised manuscript accepted for publication April
24, 2006.
Address correspondence to Muharrem Tola, MD,
Department of Radiology, Türkiye Yüksek Ihtisas
Hospital, Kızılay Sokak 4, Sihhiye-Ankara 06100,
Turkey.
E-mail: mtola@tyih.gov.tr
Muharrem Tola, MD, Mehmet Yurdakul, MD
Article
Objective. The purpose of this study was to compare velocity measurements obtained with 2 fixed
insonation angles and to investigate whether there is a difference in their ability in determining internal
carotid artery (ICA) stenosis. Methods. Eighty-seven patients with ICA stenosis were examined
with color duplex ultrasonography. Velocity measurements were made at 60° and 45° insonation
angles, and they were compared with Bland-Altman and receiver operating characteristic curve analysis.
Results. Peak systolic velocity (PSV) and end-diastolic velocity measurements obtained at the 60°
insonation angle were higher compared with those obtained at the 45° insonation angle (24.2% and
24.7%, respectively). The ICA-to-common carotid artery PSV ratio, conversely, was slightly higher
(3.9%). Although the threshold values for the same velocity parameters obtained at 2 different
insonation angles were different, the accuracy ratios (sensitivity and specificity) were not. With application
of the Society of Radiologists in Ultrasound consensus criteria to the data obtained at either of
the 2 insonation angles, the accuracy ratios of PSV and end-diastolic velocity were found to be statistically
different. In the ICA-to-common carotid artery PSV ratio, however, there were no statistically significant
differences in the accuracy ratios. Conclusions. Doppler velocity measurements made at
different fixed insonation angles show considerable differences. In determining ICA stenosis, although
optimal thresholds are different, the diagnostic performance is not different. In determining ICA stenosis
with color duplex ultrasonography, angle-specific thresholds must be determined, and examinations
must be made at a fixed angle. Key words: color duplex ultrasonography; Doppler angle;
internal carotid artery; stenosis.
olor duplex ultrasonography (CDU) has become
a standard noninvasive test for evaluation of
extracranial carotid artery stenosis. Velocity
measurements at the site of internal carotid
artery (ICA) stenosis are the primary diagnostic criteria in
CDU. In velocity measurements, there is no consensus
about the use of an insonation angle. 1 Although in some
centers, a fixed 60° angle is being used, in other centers, a
60° or smaller than 60° angle is being used. Because the
error in velocity measurements made at angles of greater
than 60° is larger, angles of greater than 60° are not used.
In determining velocity, it is assumed that any angle
between the transducer beam and the flow direction can
be compensated with angle correction. However, it has
been shown that linear array transducers used in CDU
examinations of carotid arteries measure velocity differently
depending on insonation angle. 2–8 C
These differing
measurements may cause a substantial effect on the
treatment and outcome of patients with carotid stenosis.
© 2006 by the American Institute of Ultrasound in Medicine • J Ultrasound Med 2006; 25:1187–1192 • 0278-4297/06/$3.50

Doppler Angle in Internal Carotid Artery Stenosis
1188
The purpose of this study was to compare
velocity measurements obtained with 2 fixed
insonation angles and to investigate whether
there is a difference in their ability in determining
ICA stenosis.
Materials and Methods
Eighty-seven consecutive patients who were sent
for digital subtraction angiography (DSA) were
examined with CDU before DSA. There were 60
men and 27 women with a mean age of 65 years
(range, 32–84 years). Patients were excluded from
this study for the following reasons: (1) if they
had extensive calcification to obscure the ultrasonic
signal in the stenotic area; and (2) if there
was a situation for which duplex velocity measurement
had no role (eg, occlusion and the
string sign). The string sign is very high-grade
carotid stenosis with long narrowing of the poststenotic
ICA. All patients gave their oral informed
consent, and the Institutional Review Board
approved the study.
Color duplex ultrasonography was performed
with a LOGIQ 700 system (GE Healthcare,
Milwaukee, WI) equipped with a 5- to 10-MHz linear
array transducer. The common carotid artery
(CCA) and ICA were scanned in transverse and
longitudinal planes with the B-mode and color
mode. Velocity waveforms were obtained routinely
from the CCA in the center stream approximately
2 cm below the bifurcation. The ICA was
sampled proximally just beyond the bulb widening.
When color flow imaging showed areas of
abnormal flow, which appeared as heterogeneous
color patterns, luminal narrowing, or both, the
sample volume was moved slowly from proximal
to distal in the ICA stenosis to obtain the highest
flow velocity. To bring out the change resulting
from the insonation angle more clearly, velocity
measurements were made at 60°, which is the
maximum angle suggested, and at 45°, which is
the smallest angle found to be applicable in many
carotid arteries (95%) in a preliminary study. The
highest peak systolic velocity (PSV) and end-diastolic
velocity (EDV) of blood flow in the CCA and
ICA obtained at 45º and 60º were recorded. On the
basis of these values, the ICA-to-CCA PSV ratio
(PSVr) was calculated and recorded.
Digital subtraction angiography was performed
with a Polytron V 1000 angiographic unit (Siemens
AG, Erlangen, Germany) by the Seldinger technique.
The degree of stenosis was assessed by
comparing the maximum stenotic area of the ICA
with a more normal distal portion of the depicted
ICA by North American Symptomatic Carotid
Endarterectomy Trial methods. 9 Stenosis was calculated
as [1 – (s/n)] × 100, where s was the diameter
of the maximum stenotic lumen, and n was
the diameter of the normal vessel.
The agreement between the measurements at
45° and 60° insonation angles was assessed by
the Bland-Altman method. 10,11 The following
steps were made: (1) scatterplot of the measurement
at a 45° insonation angle against the measurement
at a 60° insonation angle with the line
of equality; (2) scatterplot of the percentage of
difference between 2 measurements against
their mean; and (3) calculation of the percentage
of mean difference and 95% confidence interval
(CI) of the percentage of mean difference. The
percentage difference between 2 measurements
was calculated as (V 1 – V 2 )/(0.5 × [V 1 + V 2 ]), where
V 1 and V 2 were the measurements at 60° and 45°
insonation angles, respectively. If the percentage
of mean difference did not include 0 within 95%
CI limits, the difference between the 2 measurements
was accepted to be statistically different.
The PSV, EDV, and PSVr were evaluated by
receiver operating characteristic curve analysis in
their ability to detect ICA stenosis of 50% or
greater and 70% or greater. Receiver operating
characteristic curves based on data obtained
from CDU with 60° and 45° insonation angles
were compared by measuring the areas under the
curves (AUCs). For each Doppler parameter, the
angle-specific threshold points at which sensitivity
and specificity are optimum were determined.
After the Society of Radiologists in Ultrasound
(SRU) consensus criteria 1 (Table 1) were applied
to the data obtained at either of the 2 insonation
angles and after the sensitivity, specificity, positive
predictive value, and negative predictive
value were calculated, they were compared. The
McNemar test was used to determine any statistically
significant differences between accuracy
ratios (sensitivity and specificity).
Table 1. Society of Radiologists in Ultrasound Consensus
Velocity Criteria for Diagnosis of ICA Stenosis
Degree of Parameter
Stenosis, % ICA PSV, cm/s ICA EDV, cm/s PSVr

For statistical analysis, 2 software products were
used (MedCalc version 6.0 [MedCalc Software,
Mariakerke, Belgium] and SPSS version 11.0.0
[SPSS Inc, Chicago, IL]). P < .05 was considered to
indicate a statistically significant result.
Results
In our study, 19 carotid arteries were excluded
(12 with occlusion, 5 with extensive calcifications,
and 2 with the string sign). Velocity measurements
could be made in all arteries with the
60° insonation angle. However, in 5 ICAs and in
A
B C
11 CCAs, the 45° insonation angle could not be
obtained for velocity measurements. As a result,
for 155 arteries, 150 PSV and EDV values and 141
PSVr values obtained at the 45° insonation angle
were included for statistical evaluation. The reference
standard DSA revealed that 90 (51.7%) of the
carotid arteries had 0% to 49% stenosis; 36 (20.7%)
had 50% to 69% stenosis; 36 (20.7%) had 70% to
99% stenosis; and 12 (6.9%) were occluded.
The relationship between the measurements
obtained with both Doppler insonation angles
are shown graphically in Figure 1. Absolute
velocity measurements (PSV and EDV) obtained
with the 60° insonation angle were higher than
measurements obtained with the 45° insonation
angle. The PSVr, conversely, was slightly higher.
However, in all 3 velocity parameters, the difference
was found to be statistically significant
(Table 2).
Receiver operating characteristic curve analysis
was performed for 50% to 99% and 70% to
99% ICA stenosis on the basis of the data
obtained with 45° and 60° insonation angles on
CDU. There was no difference between the
AUCs of the same velocity parameters obtained
at the 2 different insonation angles (Tables 3
Figure 1. Scatterplots show measurement of PSV (A), EDV (B),
and PSVr (C) obtained at a 60° insonation angle versus those
obtained at a 45° insonation angle by CDU. The dashed line
indicates the equality line.
Tola and Yurdakul
J Ultrasound Med 2006; 25:1187–1192 1189

Doppler Angle in Internal Carotid Artery Stenosis
1190
Table 2. Comparison of Velocity Measurements
Obtained at 60° and 45° Insonation Angles by CDU
Mean of
Measurement Difference, % 95% CI, %
PSV 24.2 21.7–26.7
EDV 24.7 20.2–29.3
PSVr 3.9 0.5–7.4
and 4). The optimal threshold values determining
50% to 99% ICA stenosis with PSV, EDV, and
PSVr measurements were 113 cm/s, 38 cm/s, and
1.8, respectively, for the measurements obtained
at the 45° insonation angle and 137 cm/s, 42
cm/s, and 2.1 for measurements obtained at the
60° insonation angle. The optimal threshold values
determining 70% to 99% ICA stenosis with
PSV, EDV, and PSVr measurements were 153
cm/s, 51 cm/s, and 3.1 for the measurements
obtained at the 45° insonation angle and 231
cm/s, 71 cm/s, and 3.6 for measurements
obtained at the 60° insonation angle. Although
the threshold values for the same velocity
parameters obtained at 2 different insonation
angles were different, the accuracy ratios (sensitivity
and specificity) were not (Tables 3 and 4).
With application of the SRU consensus criteria
to the data obtained at either of the insonation
angles (Table 5), sensitivity and specificity of
PSVs and EDVs in determining 50% to 99% ICA
stenosis were found to be statistically different
(Table 6). In determining 70% to 99% ICA stenosis,
however, whereas sensitivity and specificity of
PSVs and sensitivity of EDVs were statistically different,
specificity of EDVs was not (Table 7). In
general, whereas sensitivity was decreasing in the
45° data, specificity was increasing. In PSVr, however,
there were no statistically significant differences
in sensitivity and specificity.
Table 3. Comparison of Diagnostic Performance of Velocity Measurements Obtained at 2 Fixed Insonation
Angles With CDU in Determining 50% to 99% ICA Stenosis
Optimum
Insonation AUC Threshold
Parameter Angle, ° (95% CI) Values Sensitivity, % Specificity, % PPV, % NPV, %
PSV 60 0.946 (0.897–0.976) 137 cm/s 89 86 82 92
45 0.951 (0.903–0.979) 113 cm/s 90 92 89 93
P = .660 P > .9 P = .063
EDV 60 0.916 (0.860–0.955) 42 cm/s 89 82 78 91
45 0.882 (0.819–0.929) 38 cm/s 79 86 80 85
P = .083 P = .07 P = .125
PSVr 60 0.942 (0.889–0.974) 2.1 91 88 84 93
45 0.955 (0.906–0.983) 1.8 93 85 82 95
P = .285 P > .9 P > .9
NPV indicates negative predictive value; and PPV, positive predictive value.
Table 4. Comparison of Diagnostic Performance of Velocity Measurements Obtained at 2 Fixed Insonation
Angles With CDU in Determining 70% to 99% ICA Stenosis
Optimum
Insonation AUC Threshold
Parameter Angle, ° (95% CI) Values Sensitivity, % Specificity, % PPV, % NPV, %
PSV 60 0.92 (0.864–0.958) 231 cm/s 87 90 69 97
45 0.92 (0.865–0.958) 153 cm/s 90 87 63 97
P = .963 P > .9 P = .125
EDV 60 0.917 (0.861–0.956) 71 cm/s 90 88 65 97
45 0.88 (0.816–0.927) 51 cm/s 87 86 60 96
P = .12 P > .9 P = .250
PSVr 60 0.903 (0.843–0.946) 3.6 87 92 73 97
45 0.927 (0.871–0.963) 3.1 89 88 66 97
P = .233 P > .9 P = .375
NPV indicates negative predictive value; and PPV, positive predictive value.
J Ultrasound Med 2006; 25:1187–1192

Discussion
Our study shows that the velocity measurements
made in carotid arteries at fixed angles of 45° and
60° are different. The fact that the optimal threshold
values created from the measurements
obtained at these 2 fixed insonation angles to
determine ICA stenosis have similar accuracy
ratios, despite their being different, reveals the
necessity that Doppler examination should also
be standardized with respect to insonation
angle, and angle-specific stenosis criteria should
be determined. All the same, further studies in
larger patient populations are warranted.
Conversion of frequency shift determined by
Doppler ultrasonography to velocity using a
Doppler equation requires manual entry of the
angle between the beam and the direction of
blood motion. Failure on the part of the operator
in determining this angle with exact accuracy
leads to faulty results. It is well known that using a
Doppler angle of greater than 60° increases the
risk of error in determining the velocity because
the error due to the cosine term in the Doppler
equation increases along with the increasing
Doppler angle. Although an error of 5° made in
measuring velocity measurement at a 45° Doppler
angle causes an error of only 9%, this error
becomes 16% at 60° and 26% at 70°. The other
source of error in connection with the Doppler
insonation angle is the intrinsic spectral broadening
(ISB) effect of linear transducers. 2,4,5 This effect
causes the actual maximal velocity to be overestimated.
The overestimation is dependent on the
insonation angle, and it increases with increasing
insonation angle. In vivo studies have also shown
that there is overestimation in maximal velocity
measurements at large angles. 3,7 Jogestrand et al 7
Table 5. Comparison of Velocity Measurements Obtained at 60° and 45° Insonation Angles With DSA Using
SRU Consensus Criteria for ICA Stenosis Diagnosis
60° Angle of Insonation 45° Angle of Insonation
Measurement DSA 0%–49% 50%–69% 70%–99% 0%–49% 50%–69% 70%–99%
PSV 0%–49% 73 15 2 83 5 0
50%–69% 1 23 10 12 17 3
70%–99% 2 2 27 2 8 20
EDV 0%–49% 69 21 0 79 9 0
50%–69% 5 25 4 12 20 0
70%–99% 2 6 23 4 13 13
PSVr 0%–49% 76 14 0 71 10 0
50%–69% 3 23 8 3 21 8
70%–99% 2 3 26 2 5 21
Table 6. Comparison of Accuracy Ratios With the Data Obtained at 45° and 60° Insonation Angles Using
the SRU Consensus Criteria in Determining 50% to 99% ICA Stenosis
Sensitivity, % Specificity, % PPV, % NPV, %
Parameter Threshold 45° 60° P 45° 60° P 45° 60° 45° 60°
PSV 125 cm/s 77 95 .001 94 81 .001 91 78 86 96
EDV 40 cm/s 74 89 .004 90 77 .0001 84 73 83 91
PSVr 2 92 92 >.9 88 84 .375 85 81 93 94
NPV indicates negative predictive value; and PPV, positive predictive value.
Table 7. Comparison of Accuracy Ratios With the Data Obtained at 45° and 60° Insonation Angles Using
the SRU Consensus Criteria in Determining 70% to 99% ICA Stenosis
Sensitivity, % Specificity, % PPV, % NPV, %
Parameter Threshold 45° 60° P 45° 60° P 45° 60° 45° 60°
PSV 230 cm/s 67 87 .031 97 90 .008 83 69 92 97
EDV 100 cm/s 43 74 .004 100 97 .125 100 85 88 94
PSVr 4 75 84 .250 93 94 >.9 72 76 94 96
NPV indicates negative predictive value; and PPV, positive predictive value.
Tola and Yurdakul
J Ultrasound Med 2006; 25:1187–1192 1191

Doppler Angle in Internal Carotid Artery Stenosis
1192
compared the measurements at 0° to 49° with those
of 50° to 62°, whereas Logason et al 3 compared the
measurements at a range of 25° to 57° with that of
fixed 60°.
Jogestrand et al 7 obtained higher accuracy results
at small Doppler insonation angles. This result supports
the fact that random operator errors made in
determining flow direction and the ISB effect are less
at small angles. In our study, however, the accuracy
ratios of the results obtained at 2 fixed insonation
angles were similar. When a fixed insonation angle is
used, the ISB effect has the same effect on all the
examinations; consequently, the inconsistencies
resulting from the use of different insonation angles
disappear when fixed angles are used. Thomas et
al 12 reported that PSVr was not affected by ISB error.
Our study, in contrary to that, showed that there was
a systematic (not random) difference, being a
smaller proportion in comparison with the other 2
absolute Doppler parameters, between the PSVr
measurements obtained at 2 fixed Doppler angles.
The answer to the question “In practice, what fixed
insonation angle should be used?” will clearly be
“any insonation angle at which the examination can
be done with ease, as long as it is less than 60°.” In
vessels such as carotid arteries, which follow the skin
surface relatively in parallel, measurements often
cannot be made at small Doppler angles. In a vessel
that lies completely parallel to the surface, examination
can be done with a 70° angle using ±20° steering.
In our study, whereas all Doppler velocity
measurements could be made at 60° with ±20° steering
and transducer tilting when necessary, measurements
of 5 (3%) ICAs and 11 (7%) CCAs could not be
made at 45°. Because measurements can easily be
made in the carotid artery at a 60° insonation angle
and also because a 60° fixed angle is widely used
nowadays, the use of a 60° insonation angle would
be appropriate for the standardization of the carotid
Doppler examination technique.
In those instances when the vascular anatomy
makes it impossible to perform an examination at
the determined insonation angle, instead of absolute
velocity measurements, it is appropriate to use
the PSVr measurement, which is less affected by IBS
error. In a case such as this, particular attention
should also be paid to the correlation of the Doppler
spectrum with gray scale and color Doppler images.
Additionally, if an examination at the determined
insonation angle is not possible, the insonation
angle at which the measurement is made should be
noted for comparison during the subsequent examination.
In conclusion, Doppler velocity measurements
made at different fixed insonation angles show considerable
differences. In determining ICA stenosis,
although optimal thresholds are different, diagnostic
performance is not different. In determining ICA
stenosis with CDU, angle-specific thresholds must
be determined, and examinations must be made at
a fixed angle.
References
1. Grant EG, Benson CB, Moneta GL, et al. Carotid artery stenosis:
gray-scale and Doppler US diagnosis—Society of Radiologists
in Ultrasound Consensus Conference. Radiology 2003;
229:340–346.
2. Daigle RJ, Stavros AT, Lee RM. Overestimation of velocity and
frequency values by multielement linear array Doppler. J Vasc
Technol 1990; 14:206–213.
3. Logason K, Barlin T, Jonsson ML, Bostrom A, Hardemark HG,
Karacagil S. The importance of Doppler angle of insonation on
differentiation between 50–69% and 70–99% carotid artery
stenosis. Eur J Vasc Endovasc Surg 2001; 21:311–313.
4. Thrush AJ, Evans DH. Intrinsic spectral broadening: a potential
cause of misdiagnosis of carotid artery disease. J Vasc Invest
1995; 1:187–192.
5. Hoskins PR. Measurement of maximum velocity using duplex
ultrasound systems. Br J Radiol 1996; 69:172–177.
6. Steinman AH, Tavakkoli J, Myers JG, Cobbold RSC, Johnston
KW. Sources of error in maximum velocity estimation using linear
phased-array Doppler systems with steady flow.
Ultrasound Med Biol 2001; 27:655–664.
7. Jogestrand T, Lindqvist M, Nowak J. Diagnostic performance
of duplex ultrasonography in the detection of high grade internal
carotid artery stenosis. Eur J Vasc Endovasc Surg 2002;
23:510–518.
8. Tahmasebpour H, Cooperberg P, Segan-Hoffman J, Normand
L, Fix C. Velocity quantifications in carotid ultrasound with
Doppler angle set at 44 vs 60 degrees. Paper presented at:
Annual Meeting of the Radiological Society of North America;
November 30, 2003; Chicago, IL.
9. North American Symptomatic Carotid Endarterectomy Trial
Collaborators. Beneficial effect of carotid endarterectomy in
symptomatic patients with high-grade carotid stenosis. N Engl
J Med 1991; 325:445–453.
10. Bland JM, Altman DG. Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet
1986; 8:307–310.
11. Bland JM, Altman DG. Measuring agreement in method comparison
studies. Stat Methods Med Res 1999; 8:135–160.
12. Thomas N, Taylor P, Padayachee S. The impact of theoretical
errors on velocity estimation and accuracy of duplex grading
of carotid stenosis. Ultrasound Med Biol 2002; 28:
191–196.
J Ultrasound Med 2006; 25:1187–1192