(Stroke. 1997;28:692-695.)
© 1997 American Heart Association, Inc.
Articles |
Use of a Decibel Threshold in Detecting Doppler Embolic Signals
From the Department of Neurology, King's College School of Medicine and Dentistry, and the Institute of Psychiatry, London, UK.
Correspondence to Dr Hugh Markus, Department of Neurology, Institute of Psychiatry, De Crespigny Park, London, SE5 8AF, UK. E-mail h.markus{at}iop.bpmf.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Background and Purpose To improve reproducibility and reliability in the identification of embolic signals detected with the use of Doppler ultrasound, many studies have used an intensity threshold. However, variable thresholds between 3 and 12 dB have been used, and often the method of measurement of intensity is not stated. Potentially different methods of measurement could result in different intensity measurements for the same embolic signal. We determined the effect of these differences using commercial transcranial Doppler systems.
Methods We analyzed 81 embolic signals recorded from the middle cerebral arteries of patients with carotid artery disease using three different methods of measuring intensity that had been previously used in research studies. In method 1 individual time frames of the frequency spectra were analyzed, in method 2 a color-coded intensity scale was used, and in method 3 automated software was used.
Results There was a highly significant correlation between measurements made by the different techniques (method 1 versus method 2: r=.68, P<.0001; method 1 versus method 3: r=.66, P<.0001; method 2 versus method 3: r=.70, P<.0001). However, the absolute values of intensity for the same embolic signals varied markedly for the different methods. For example, a 4-dB threshold according to method 1 was equivalent to an approximately 7-dB threshold measured by method 2. These differences had major effects on the proportion of embolic signals detected with the use of the same decibel threshold but with intensity measured in the different ways. For example, using a threshold of 7 dB would result in only 4.9% of signals being missed by method 2 but 42.2% and 51.4% being missed by methods 1 and 3, respectively.
Conclusions Our results demonstrate that the intensities of the same embolic signals, recorded with the same parameters, are markedly different when analyzed in the different ways used in previous studies. This has important implications when a decibel threshold is used and emphasizes that criteria developed by one investigator on one machine cannot be used by another investigator without initial reevaluation. This could account for some of the differences in frequencies of embolic signals reported in previous clinical studies.
Key Words: carotid stenosis • cerebral embolism • ultrasonics
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Cerebral embolus detection with the use of Doppler ultrasound has many potential applications in the management of patients with cerebrovascular disease. In certain conditions, such as carotid artery stenosis, the presence of embolic signals appears to correlate with indicators of disease activity.1 2 However, initial studies have produced widely varying proportions of patients in whom embolic signals can be detected. For symptomatic carotid stenosis this has varied between 20% and more than 90%.3 4 5 6 7 Reproducibility studies within one center or between two centers have shown that good interobserver reproducibility can be obtained but there may be disagreement, particularly for signals of low intensity,7 8 9 and studies between larger groups of observers have resulted in less agreement. Many researchers in this field have found that often patients with clear embolic signals also exhibit very small increases in relative intensity accompanied by a characteristic sound, and deciding which to count as definite embolic signals can be difficult. The use of a decibel threshold is one way of resolving this problem. Only signals above a certain intensity will then be recorded as embolic signals. This results in greater interobserver agreement since disagreement usually occurs for low-intensity signals.8 Many recent studies have included the use of a decibel threshold, which has varied from 3 to 12 dB.7 9 10 11 12 13 14 15 16 17 However, few of these state how the intensity was calculated. A number of studies have been performed in which an intensity threshold suggested by one investigator on one machine with certain settings is applied to either a different machine or to the same machine with different settings, without any revalidation.
Intensity is calculated from the logarithm of the ratio of the power of the embolic signal to that of the Doppler spectrum in the absence of any embolic signal. However, each of these two parameters can be measured in a number of ways. We determined whether three different methods employed in clinical studies in which commercially available transcranial Doppler equipment was used resulted in significant differences in the analysis of embolic signals from patients with carotid artery disease. We specifically chose this patient group because the embolic signals in patients with carotid stenosis are of lower intensity than those detected in patients with prosthetic heart valves18 or during cardiopulmonary bypass and present greater diagnostic difficulty.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Recordings were made from the middle cerebral artery ipsilateral to a symptomatic carotid stenosis in eight patients. Recordings were made on an EME TC 2000 transcranial Doppler machine with a 2-MHz transducer, a sample volume of 10 mm, and a depth of 45 to 52 mm. The recordings had previously been blindly evaluated by four observers from two centers in an interobserver reproducibility study.8 Eighty-one embolic signals that had been detected by all four observers were used for the subsequent analysis. One method (method 3 below) of determining the relative intensity increase of an embolic signal required that the embolic signal was detected with the use of an automated detection system. Even when a low detection threshold was used, 9 embolic signals were not detected with this method; therefore, only 72 are included in analyses in which this method is used.
The Doppler audio signal was recorded on digital audiotape before any fast Fourier transform processing and subsequently played back into two transcranial Doppler systems: (1) the same EME TC2000 and (2) an EME Pioneer 4/40 system. The intensity of each Doppler embolic signal was determined by three methods used in previously published studies, as described below.
Method 1
Individual time frames of the fast Fourier transform were
analyzed on the TC2000 according to a previously described
method.19 The maximum relative power amplitude (RPA)
associated with the embolic signal was recorded. The background RPA
in the absence of an embolic signal was measured from a Doppler
spectrum of the previous or next cardiac cycle, at the same point in
the cycle and at the same velocity. A mean of three background readings
was taken. Intensity was then calculated from the following equation:
Intensity Increase=10 log (Maximum RPA of Embolic Signal/RPA in Absence
of Embolic Signal).
Method 2
With the use of the EME Pioneer, the intensity was calculated
from the color-coded intensity scale on the screen. This can be
adjusted so that its intensity can be measured to the nearest
decibel.20 The gain was reduced until the color of the
adjacent cardiac cycle reached 0, and the intensity of the embolic
signal was then determined.
Method 3
With the use of the EME Pioneer, the intensity of each embolic
signal was calculated with the automatic embolus software supplied with
the machine. The algorithm determines the power of the embolic signal
over the whole spectral line; similarly, the background power is
calculated over the whole spectral line using a running average of
background intensity over the preceding spectral lines.
Previously it has been suggested that an appropriate intensity threshold may be determined by measuring the intensity increase occurring with random episodes of Doppler "speckle"7 ; these variations in intensity occur in the normal Doppler spectra and result from a number of factors, including nonuniformity of the ultrasound field and nonuniformity of the distribution of red blood cell scatterers. For this reason, the intensity increases associated with 200 episodes of random Doppler speckle using recordings from normal volunteers, made at the same depth, sample volume, and gain, were calculated with intensity measured as in methods 1 and 2. It was not possible to perform this analysis with the automated embolic signal detection software used in method 3.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The relationship between embolic signal intensity measured by the different methods is demonstrated in Fig 1
. There was a
highly significant correlation between measurements made by the
different techniques (method 1 versus method 2: r=.68,
P<.0001; method 1 versus method 3: r=.66,
P<.0001; method 2 versus method 3: r=.70,
P<.0001). However, as can be seen in Fig 1, the absolute
values of intensity for the same embolic signals varied markedly for
the different methods, with intensity usually being higher for method
2. A 4-dB threshold according to method 1 was equivalent to an
approximately 7-dB threshold measured by method 2. We estimated the
effect of these differences on the proportion of embolic signals
detected using the same decibel threshold but with intensity measured
in different ways (Table). The percentage of embolic
signals detected varied markedly; for example, using a threshold of 7
dB would result in only 4.9% of signals being missed by method 2 but
42.2% and 51.4% being missed by methods 1 and 3, respectively. A
similar difference was apparent when an appropriate threshold was
determined from analysis of the relative intensity increase
associated with episodes of Doppler speckle. The use of an upper
limit of 2 SD above the mean resulted in a threshold of greater than 4
dB for method 1 and greater than 7 dB for method 2 (Fig 2).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our results demonstrate that the intensity of the same embolic signals, recorded with the use of the same parameters, can be markedly different when analyzed in different ways. This has important implications when a decibel threshold is used and emphasizes that criteria developed by one investigator on one machine cannot be used by another investigator without initial reevaluation. Such use can lead to great differences in the number of detected embolic signals. For example, the use of a 7-dB threshold measured by method 2 resulted in 95% detection of embolic signals, whereas the use of the same threshold with intensity measured with method 1 or 3 would result in approximately half of the embolic signals being missed. The figures may be even more discrepant if very-low-intensity embolic signals are used. We only studied embolic signals that four observers had agreed were present, and this by definition introduced a detection threshold. Furthermore, using an inappropriately low threshold for the method of measurement of intensity may lead to many episodes of Doppler speckle being inappropriately counted as embolic signals. This may account for the detection of "embolic signals" in 100% of patients in a recent study when a low-intensity threshold was used.21
The differences between the results obtained by the three methods may have a number of explanations. The power increase associated with the embolic signal can be measured in a number of ways. These include the peak increase at one velocity, the area under the power increase measured both across velocities and across time, or the power increase along the whole spectral line, which will include the power increase of the embolic signal and also of the background Doppler spectrum at other velocities. Similarly, the background power may be measured in a number of ways. The background power can be measured at the same velocity or at all velocities, at the same point of the cardiac cycle or averaging across the whole cardiac cycle, and only within the Doppler spectrum or along the whole spectral line. For example, the background intensity is higher in diastole than in systole, and therefore the position of the cardiac cycle will alter measurements. Similarly, if the whole spectral line is used, for technically poor recordings with artifactual extraspectral noise, the background power will appear higher, resulting in a lower intensity increase of the embolic signal.
In this study we used the same recordings to measure intensity in three ways. In practice there are additional problems with comparing intensity measurements from one study to another study. The intensity of an embolic signal will be highly dependent on the recording parameters used. For example, the shorter the length of the sample volume, the greater will be the ratio of the power of the embolic signal to the background spectra and therefore the higher the intensity of the embolic signal. The intensity may vary if the depth of insonation is varied because of differences in sample volume width for focused ultrasound beams at different depths. Differences may arise even with the same apparatus and settings if, for example, the middle cerebral artery lies in the center of the sample volume in one patient, but if it is not centrally insonated in another patient. Additional problems may arise if the degree of fast Fourier transform overlap is inadequate.22 The intensity increase will be greater if the embolic signal is processed in the middle of one time window rather than falling between two adjacent time windows. Nevertheless, in practice these differences can be reduced by using standardized settings and recording parameters, as in this study. Some of these difficulties suggest that the use of a decibel threshold is inappropriate. However, different investigators have different thresholds at which they interpret a small intensity increase as an embolic signal, and therefore we believe that an intensity threshold may aid in comparing data from different studies. We suggest that an approximate threshold can be determined by analyzing the intensity increases resulting from random Doppler speckle, as in this study, but that this should be performed by each center on its equipment and at its usual settings, and the equipment setting should then be kept constant for subsequent recordings. An additional way of ensuring the use of an appropriate intensity threshold is to analyze mixed data from patients and normal control subjects while blinded to the source of an individual recording. Embolic signal detection depends on features other than intensity increase alone, such as the unidirectional velocity of the signal and the characteristic sound. Further technological advances such as the multigate technique23 24 may allow more unambiguous detection of embolic signals and determination of which low-intensity signals result from circulating emboli.
The marked difference in intensity measurements with the use of commercially available transcranial Doppler machines has important implications for clinical studies of embolic signal detection. If an intensity threshold is to be established, this should be determined by each center using its own equipment. Wherever possible, all recordings should then be made at a similar depth with a similar sample volume and settings. These findings also have important implications if intensity measurements are to be used to gain information about the composition of the underlying embolic material.
|
Acknowledgments |
|---|
This study was supported by a British Heart Foundation project grant.
Received October 14, 1996; revision received December 31, 1996; accepted December 31, 1996.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1. Markus HS, Harrison MJ. Microembolic signal detection using ultrasound. Stroke. 1995;26:1517-1519.
2.
Siebler M, Nachtmann A, Sitzer M, Rose G, Kleinschmidt
A, Rademacher J, Steinmetz H. Cerebral microembolism and the
risk of ischemia in asymptomatic high-grade
internal carotid artery ischemia. Stroke. 1995;26:2184-2186.
3. Georgiadis D, Grosset DG, Quin RO, Nichol JA, Bone I, Lees KR. Detection of intracranial emboli in patients with carotid disease. Eur J Vasc Surg. 1994;8:309-314. [Medline] [Order article via Infotrieve]
4.
Valton L, Larrue V, Arrue P, Geraud G, Bes A.
Asymptomatic cerebral embolic signals in patients with
carotid stenosis: correlation with the appearance of plaque
ulceration on angiography. Stroke. 1995;26:813-815.
5.
Siebler M, Kleinschmidt A, Sitzer M, Steinmetz H,
Freund HJ. Cerebral microembolism in symptomatic and
asymptomatic high-grade internal carotid artery
stenosis. Neurology. 1994;44:615-618.
6. Babikian VL, Hyde C, Pochay V, Winter MR. Clinical correlates of high-intensity transient signals detected on transcranial Doppler sonography in patients with cerebrovascular disease. Stroke. 1994;25:1570-1573. [Abstract]
7.
Markus HS, Thomson N, Brown MM.
Asymptomatic cerebral embolic signals in
symptomatic and asymptomatic carotid artery
disease. Brain. 1995;118:1005-1011.
8.
Markus HS, Bland JM, Rose G, Sitzer M, Siebler
M. How good is inter-center agreement in the identification of
embolic signals in carotid artery disease?. Stroke. 1996;27:1249-1252.
9.
Siebler M, Sitzer M, Rose G, Bendfeldt D, Steinmetz
H. Silent cerebral embolism caused by neurologically
symptomatic high-grade carotid stenosis: event
rates before and after carotid
endarterectomy. Brain. 1993;116:1005-1015.
10.
Forteza AM, Babikian VL, Hyde C, Winter M, Pochay
V. Effect of time and cerebrovascular symptoms on the prevalence
of microembolic signals in patients with cervical
carotid stenosis. Stroke. 1996;27:687-690.
11.
Nabavi DG, Georgiadis D, Mumme T, Schmid C, Mackay TG,
Scheld HH, Ringlestein EB. Clinical relevance of intracranial
microembolic signals in patients with left
ventricular assist devices. Stroke. 1996;27:891-896.
12. Eicke BM, Barth V, Kukowski B, Werner G, Paulus W. Cardiac microembolism: prevalence and clinical outcome. J Neurol Sci. 1996;136:143-147. [Medline] [Order article via Infotrieve]
13. Droste DW, Decker W, Siemens H, Kaps M, Schulte-Altedorneburg G. Variability in occurrence of embolic signals in long term transcranial Doppler recordings. Neurol Res. 1996;18:25-30. [Medline] [Order article via Infotrieve]
14. Grosset DG, Georgiadis D, Abdullah I, Bone I, Lees KR. Doppler emboli signals vary according to stroke subtype. Stroke. 1994;25:382-384. [Abstract]
15.
Braekken SK, Russell D, Brucher R, Svennevig J.
Incidence and frequency of cerebral embolic signals in patients with a
similar bileaflet mechanical heart valve. Stroke. 1995;26:1225-1230.
16.
Van Zuilen EV, Moll FL, Vermeulen FE, Mauser HW, van
Gijn J, Ackerstaff RG. Detection of cerebral microemboli by
means of transcranial Doppler monitoring before and
after carotid endarterectomy.
Stroke. 1995;26:210-213.
17.
Sliwka U, Job F, Wissuwa D, Diehl RR, Flachskampf F,
Hanrath P, Noth J. Occurrence of transcranial
Doppler high-intensity transient signals in patients with potential
cardiac sources of embolism. Stroke. 1995;26:2067-2070.
18.
Grosset DG, Georgiadis D, Kelman AW, Lees KR.
Quantification of ultrasound emboli signals in patients with cardiac
and carotid disease. Stroke. 1993;24:1922-1924.
19.
Markus H, Loh A, Brown MM. Computerized
detection of cerebral emboli and discrimination from artifact using
Doppler ultrasound. Stroke. 1993;24:1667-1672.
20.
Russell D, Madden KP, Clark WM, Sandset PM, Zivin
JA. Detection of arterial emboli using Doppler
ultrasound in rabbits. Stroke. 1991;22:253-258.
21. Demarin V, Rundek T, Miletic B. Prevalence of cerebral emboli in patients with carotid artery disease. Cerebrovasc Dis. 1996;6(suppl 3):243. Abstract.
22.
Markus HS. Importance of time window overlap in
the detection and analysis of embolic signals.
Stroke. 1995;26:2044-2047.
23.
Georgiadis D, Gocke J, Hill M, Konig M, Nabavi DG,
Stoghauer F, Zunker P, Ringlestein EB. A novel technique for
identification of Doppler microembolic signals
based on the coincidence method. Stroke. 1996;27:683-686.
24.
Molloy J, Markus HS. Multigated Doppler
ultrasound in the detection of emboli in a flow model and embolic
signals in patients. Stroke. 1996;27:1548-1552.
This article has been cited by other articles:
 |
|
 |
|
  V. K. Sharma, G. Tsivgoulis, A. Y. Lao, M. D. Malkoff, A. W. Alexandrov, and A. V. Alexandrov Quantification of Microspheres Appearance in Brain Vessels: Implications for Residual Flow Velocity Measurements, Dose Calculations, and Potential Drug Delivery Stroke, May 1, 2008; 39(5): 1476 - 1481. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Markus and A. MacKinnon Asymptomatic Embolization Detected by Doppler Ultrasound Predicts Stroke Risk in Symptomatic Carotid Artery Stenosis Stroke, May 1, 2005; 36(5): 971 - 975. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Plochl, C. G. Krenn, D. J. Cook, E. Gollob, T. Pezawas, H. Schima, O. Ipsiroglu, G. Wollenek, and G. Grubhofer Can hypocapnia reduce cerebral embolization during cardiopulmonary bypass? Ann. Thorac. Surg., September 1, 2001; 72(3): 845 - 849. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cullinane, G. Reid, R. Dittrich, Z. Kaposzta, R. Ackerstaff, V. Babikian, D. W. Droste, D. Grossett, M. Siebler, L. Valton, et al. Evaluation of New Online Automated Embolic Signal Detection Algorithm, Including Comparison With Panel of International Experts Stroke, June 1, 2000; 31(6): 1335 - 1341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Markus, M. Cullinane, and G. Reid Improved Automated Detection of Embolic Signals Using a Novel Frequency Filtering Approach Stroke, August 1, 1999; 30(8): 1610 - 1615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Taylor, M. A. Borger, R. D. Weisel, L. Fedorko, and C. M. Feindel Cerebral microemboli during cardiopulmonary bypass: increased emboli during perfusionist interventions Ann. Thorac. Surg., July 1, 1999; 68(1): 89 - 93. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Molloy and H. S. Markus Asymptomatic Embolization Predicts Stroke and TIA Risk in Patients With Carotid Artery Stenosis Stroke, July 1, 1999; 30(7): 1440 - 1443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Molloy, J. F. Martin, P. A. Baskerville, S. C. A. Fraser, and H. S. Markus S-Nitrosoglutathione Reduces the Rate of Embolization in Humans Circulation, October 6, 1998; 98(14): 1372 - 1375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Cullinane, R. Wainwright, A. Brown, M. Monaghan, and H. S. Markus Asymptomatic Embolization in Subjects With Atrial Fibrillation Not Taking Anticoagulants : A Prospective Study Stroke, September 1, 1998; 29(9): 1810 - 1815. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. M. Keunen, C. J. Stam, D. L. J. Tavy, W. H. Mess, B. M. Titulaer, and R. G. A. Ackerstaff Preliminary Report of Detecting Microembolic Signals in Transcranial Doppler Time Series With Nonlinear Forecasting Stroke, August 1, 1998; 29(8): 1638 - 1643. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Bernd Ringelstein, D. W. Droste, V. L. Babikian, D. H. Evans, D. G. Grosset, M. Kaps, H. S. Markus, D. Russell, and M. Siebler Consensus on Microembolus Detection by TCD Stroke, March 1, 1998; 29(3): 725 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Markus, R. Ackerstaff, V. Babikian, C. Bladin, D. Droste, D. Grosset, C. Levi, D. Russell, M. Siebler, and C. Tegeler Intercenter Agreement in Reading Doppler Embolic Signals : A Multicenter International Study Stroke, July 1, 1997; 28(7): 1307 - 1310. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||