Cerebral Emboli Monitoring Using Transcranial Doppler Ultrasonography in Adults and Children: A Review of the Current Technology and Clinical Applications in the Perioperative and Intensive Care Setting : Anesthesia & Analgesia

Secondary Logo

Journal Logo

Review Articles: Narrative Review Article

Cerebral Emboli Monitoring Using Transcranial Doppler Ultrasonography in Adults and Children: A Review of the Current Technology and Clinical Applications in the Perioperative and Intensive Care Setting

Kussman, Barry D. MBBCh, FFA(SA)*,†; Imaduddin, Syed M. SM; Gharedaghi, Mohammad Hadi MD, MPH*,†; Heldt, Thomas PhD; LaRovere, Kerri MD§,‖

Author Information
Anesthesia & Analgesia 133(2):p 379-392, August 2021. | DOI: 10.1213/ANE.0000000000005417
  • Free
  • SDC
  • CME Test
  • Continuing Medical Education


Cerebral embolism is a major mechanism of transient ischemic attack (TIA), stroke, and silent brain infarcts (SBI).1 Doppler ultrasound for the detection of gas emboli was first used in the 1960s as an early warning of decompression sickness in divers.2 Spencer et al3 reported the use of “ultrasonics” in the determination of arterial aeroembolism during open-heart surgery in 1969. The sensors in this study were placed in-line on the extracorporeal circuit and around the aortic arch and its major branches. In 1979, investigators at the Academy of Medical Sciences in Moscow demonstrated the ability to evaluate cerebral hemodynamics in premature infants with ultrasound by measurement of cerebral blood flow velocity (CBFV) through the anterior fontanelle.4 The landmark study by Aaslid et al5 in 1982 established that it was possible to measure CBFVs through the adult skull by using low ultrasound frequencies (2 MHz). Thereafter, transcranial Doppler (TCD) sonography was used clinically as a noninvasive, bedside technique for the measurement of CBFV as a surrogate of cerebral blood flow (CBF) and as a method for detection of cerebral embolism.

CBFV has shown varying degrees of correlation with absolute CBF and changes in CBF measured with invasive and noninvasive techniques.6–12 Current clinical applications for measurement of CBFV include evaluation of cerebrovascular physiology (syncope and cardiac arrest, directions of flow through the circle of Willis, vascular reactivity, autoregulation, responses to clinical interventions, sleep physiology), in cerebrovascular disease (patency of vessels, vasospasm following subarachnoid hemorrhage, sickle cell disease), carotid endarterectomy (CEA; hypoperfusion with clamping, postoperative hyperperfusion), cardiac surgery (adequacy of cardiopulmonary bypass [CPB] flow rates), cardiac catheterization (cardiac output changes with interventions), and monitoring in the intensive care unit (traumatic brain injury, diagnosis of brain death, intraventricular hemorrhage, detection of increased intracranial pressure). These indications are beyond the scope of this review and are described in authoritative texts and reviews for the interested reader.13–16

The aim of this review is to present an update on the current status of TCD sonography for the evaluation of cerebral embolism in the perioperative and intensive care setting. We will review basic ultrasound physics, the technical performance, and principles of TCD pertaining to emboli monitoring, the clinical applications, and future directions.


Measurement of CBFV with TCD ultrasound may use continuous or pulsed insonation. The former technique is based on the Doppler principle in which sinusoidal ultrasound waves are continuously generated and are scattered off moving red blood cells (RBCs) producing a velocity-dependent shift (fD) in the ultrasound frequency (Doppler shift)

fD= fRfT=2 fTvc cosθ,

where fR and fT are the frequencies of the received and transmitted ultrasound signals, respectively, v is the velocity of the moving RBCs, c is the speed of sound in the tissue, θ is the angle between the incident ultrasound beam and the velocity vector of the RBCs (Doppler angle). The negative sign indicates that the received frequency is smaller than the transmitted frequency when flow is directed away from the source.17 Instead of continuous insonation, short pulses (~5 microseconds for a 10-cycle pulse at 2 MHz) may be used in applications where judging the depth of the target vessel is also important. The backscattered echoes for each pulse are recorded and the time of arrival is used to gauge the depth of the scattering particles. The pulses are produced at fixed pulse repetition intervals (PRIs) that may vary between 0.1 and 0.2 milliseconds corresponding to pulse repetition frequencies (PRFs) (PRF = 1/PRI) between 5 and 10 kHz. Shorter pulses lead to improved depth resolution (and a smaller sample volume) at the expense of lower signal to noise ratio.18 The movement of the scattering particles leads to an interpulse sinusoidal variation in the signal received from the target depth. The frequency of this sinusoidal variation (Doppler shift) can be shown to be that of Equation 1. In practice, a spectrogram of the received backscattered signals is computed to determine the frequency shift. This computation typically involves applying the fast Fourier transform (FFT) algorithm to overlapping temporal segments of the received signal. While scattering occurs from groups of cells moving at different velocities, the CBFV reported in clinical TCD examinations is, by convention, the envelope of the spectrogram corresponding to the maximum red cell velocity in the sample volume at a given time.

The different reflectivity (acoustic impedance) of emboli compared to surrounding blood is the basis for the detection of cerebral emboli. In other words, the intensity of the returning Doppler signal from microemboli is greater than that from surrounding RBCs.19 These transient increases in signal intensity over that of blood (relative intensity increase) are referred to as high-intensity transient signals (HITS) that may represent emboli composed of gaseous or solid (particulate) matter, or reflect artifact.19 HITS are discussed in more detail later.

Table 1. - Consensus for Reporting of Emboli Studies
1. Ultrasound device
2. Transducer type and size
3. Insonated artery
4. Insonation depth
5. Algorithms for signal intensity measurement
6. Scale settings
7. Detection threshold
8. Axial extension of sample volume
9. FFT size (number of points used)
10. FFT length (time)
11. FFT overlap
12. Transmitted ultrasound frequency
13. High-pass filter settings
14. Recording time
Reproduced with permission from Ringelstein et al.20
Abbreviation: FFT, fast Fourier transform.

The settings of the ultrasound system can have significant effects on HITS detection. Important technical parameters in this regard are the transmitted ultrasound frequency (fT), PRF, sample volume, FFT frequency resolution and temporal overlap, dynamic range of the instrumentation, filter settings, the embolus-to-blood ratio (EBR) detection threshold, and the recording time.20 An expert consensus committee recommended keeping the technical parameters listed in Table 1 constant during recordings performed for clinical care and research to allow for comparison between studies.20


Transcranial Windows

The temporal and occipital regions of the skull allow adequate transmission of ultrasound waves in the majority of patients and are the most commonly used sites for access to the anterior and posterior circulations around the Circle of Willis. In neonates and infants, the anterior circulation may also be insonated through an open anterior fontanelle.21

Interrogation of the cerebral arteries through the temporal window (thinnest part of the skull) is not possible in approximately 8% of adults.22 The thickness and texture of the temporal bone, as can be determined by computed tomography (CT), are important factors for temporal acoustic window failure.23 In adults, the middle layer (diploe) of cancellous bone in the skull significantly contributes to loss (absorption and reflection) of the ultrasound signal.24 An infant skull, on the other hand, has no diploe layer, and accordingly, there is very little transmission loss of signal.24

An inadequate occipital window occurs in around 9% of adults. Since the occipital window (also referred to as the transforaminal window) is an anatomical foramen (the foramen magnum) covered by tendons and aponeurosis, ultrasonic wave attenuation here is primarily caused by absorption as opposed to scattering. In adults, female sex and older age (osteoporosis causes ultrasound wave scattering) seem to be associated with inability to obtain TCD signals from the temporal window, whereas male sex (larger and thicker bones and tendons) have higher failure rates from the occipital window.23

Angle of Insonation

The angle of insonation (AOI) is the angle between the ultrasound beam and the blood cell’s velocity vector and is dependent on a vessel’s orientation and on the operator. It is important that the AOI be as close to 0 as possible (cosine of 0° is 1) to maximize the accuracy of the measured velocity. When the AOI ranges from 0° to ±30° the maximum absolute error is <15%.5 AOIs >30° results in a progressive reduction in calculated CBFV amplitudes, and CBFV cannot be measured when the ultrasound probe is oriented at 90° to the direction of blood flow. Since clinically significant or pathologic emboli would be expected to reduce cerebral perfusion, it is important to avoid falsely attributing low CBFVs to pathologic emboli when the AOI is not the lowest achievable. Although transcranial color-coded duplex ultrasonography could measure the AOI and allow for angle correction, this is not possible with TCD monitoring systems. However, the small size of the transtemporal window usually limits the AOI of the middle cerebral artery (MCA) to <30°.

Operator Dependency

Performance of TCD is highly operator dependent. To obtain reliable measurements, studies should be performed by persons with experience in TCD examinations and signal optimization should routinely be used to ensure high-quality data acquisition.25,26 Embolic signal interpretation is also operator dependent. Agreement between automatic embolus detection and the human observer was found to be poor (κ = 0.24), whereas inter- and intraobserver reliability was high (κ = 0.72–0.79).27 The probability of agreement between centers using a 7 dB threshold is 0.9.28 Emboli detection currently relies on the expert human observer to visually inspect the Doppler waveform offline and differentiate HITS into true emboli versus artifact. This process is exceedingly time-consuming and labor intensive. It is also important to note that there is no ground truth. Ground truth refers to knowledge of the composition and size of the emboli, something that can be known and controlled in laboratory studies (eg, glass or plastic beads, big bubbles) albeit not very realistic. However, in the clinical setting, ground truth determination of emboli composition and size is currently not possible.

Maintaining Probe Position

Emboli detection requires prolonged TCD monitoring, particularly if emboli are infrequent, to allow adequate time to detect possible emboli, and maintaining uninterrupted signal acquisition for minutes to hours is challenging.29 Different headframes have been designed to hold the ultrasound probe in place and to minimize motion artifacts resulting from head or probe movement. Available pediatric head frames are too large, however, for the very young (particularly <6 years when head size varies). In the perioperative setting, the proximity of the head frame to the surgical field and/or restricted space at the head of the bed increase the likelihood of probe dislodgement and loss of signal. Probe fixation is challenging in children on extracorporeal membrane oxygenation (ECMO) and cannulated via the carotid artery and internal jugular vein. Placement of a headframe or rotation of the head for bilateral emboli monitoring could lead displacement of the ECMO cannulae.22


Physical Characteristics of Emboli

Table 2. - Important Diameters (in Micrometers) for Embolus Detection
1800–4000 Middle cerebral artery
400–1100 Anterior cerebral artery
600 Smallest platelet embolus detected with TCD monitoring (rabbit aorta)a
500 Smallest fat or atheroma emboli detected with TCD monitoring (rabbit aorta)a
300–800 Lenticulostriate arteries occluded with lacunar infarcts, some caused by emboli
300 Smallest air embolus detectable with TCD monitoring (rabbit aorta)a
150–200+ Size of microemboli that preferentially go to the watershed area of the MCA distribution (human cadaver with glass microspheres)
193 Polystyrene microsphere clearly detectable with 5-MHz CW Doppler instrument and automated monitoring of change in Doppler intensity (in vitro with plastic tubing)
150 Smallest detectable static bubble with 2- to 5-MHz CW Doppler instruments (in vitro)
130 Largest air bubble that will cross an intact dog lung following pretreatment with aminophyllineb 100–250
100–250 Diameter of visible retinal arterioles measured with an ophthalmoscope
115 Smallest reliably detected single plastic sphere (in vitro)
70 Smallest detectable moving single air bubble with 2- to 5- MHz CW Doppler instrument (in vitro)
50 Smallest detectable plastic sphere (in vitro, intermittent detection)
40 Largest observed vessels in arteriolar dilations following open-heart surgery
30 Smallest moving plastic spheres detected with TCD monitoring (presented as a cloud in vitro)
25 Smallest detectable retinal microvessel with fluorescein angiography
Pore size for arterial line filters in open-heart surgery
14 Largest emboli that will pass through an intact dog lunga
10 Diameter of white blood cell
Hypothesized smallest air or solid embolus that as an isolated event would produce an infarct
Smallest gaseous bubble that will remain stable in blood without collapsing
7 Diameter of red blood cell
5 Smallest observed vessels with arteriolar dilations following open-heart surgery
2–6 Size of air emboli that freely pass through rat and rabbit lung
Size of air emboli that, when presented in a concentrated solution, produce
Doppler signal enhancement rather than emboli blips
2 Size of air emboli able to pass through a damaged blood-brain barrier (rat glioma model)
Vessel sizes are in adults. Reproduced with permission from Babikian and Wechsler.35
Abbreviations: CW, continuous wave; MCA, middle cerebral artery; TCD, transcranial Doppler.aSmaller sizes were not tested; therefore, the smallest detectable size is not known.bLarger sizes were not tested; therefore, the largest size that will pass through the lungs is not known.

Emboli may be gaseous (air, carbon dioxide) or solid (atheromatous plaque, platelet aggregates, thrombus, lipid droplets, or foreign material). Macroemboli (>100 μm) can produce a clinically detectable TIA or stroke by occluding larger arteries that supply focal vascular territories. In contrast, microemboli (<100 μm) generally occlude only small arterioles and capillaries. Emboli also produce endothelial changes that disrupt the blood-brain barrier.30,31 The composition of emboli is frequently condition-specific (eg, CEA [plaque], intracardiac thrombus [clot], fat embolism syndrome [fat globules]).32–34 Important diameters for embolus detection are presented in Table 2; the vessel sizes are for adults.35

High-Intensity Transient Signals

HITS representing true emboli appear in the Doppler spectrum as a result of transient increases in the ratio of the acoustic power backscattered from the embolus relative to that of the surrounding flowing blood.20 This is referred to as the EBR and is defined as the ratio of reflected power from an embolus normalized by the power calculated over sample volumes not containing any emboli.36 Criteria for the basic identification of emboli by TCD were developed by a consensus committee in 1995 and include brief duration (<300 milliseconds); primarily unidirectional propagation (as determined from the FFT); high intensity (>3 dB above the background); and accompanied by an audible component (chirp, pop).32 HITS can occur as single events, as clusters (defined as >10 HITS per 3–5 cardiac cycles) or with a “curtain effect” (per 3–5 cardiac cycles) (Figure 1, Supplemental Digital Contents 1–3, Videos 1A–C, https://links.lww.com/AA/D392, https://links.lww.com/AA/D393, https://links.lww.com/AA/D394). An embolic signature is also evident in the time domain signal as a rhomboid or hexagonal envelope with frequency and/or amplitude modulation (Figure 2).20

Figure 1.:
Categorization of HITS. A, Single HITS. B, HITS cluster (>10 HITS per 3–5 cardiac cycles). C, HITS with “curtain effect” per 3–5 cardiac cycles. HITS within the Doppler spectrum are indicated by the yellow vertical lines with higher power than the flowing blood. HITS indicates high-intensity transient signals; PSD, power spectral density.
Figure 2.:
Sinusoidal correlation (quadrature) accompanying a HITS. The correlation appears as a rhomboid envelope in the time domain with frequency and amplitude modulation. HITS indicates high-intensity transient signals.
Figure 3.:
Embolic signatures in (a) the middle cerebral artery and (b) the anterior cerebral artery. The spectrograms with the HITS are restricted to the point in time that the PMD embolic track crosses the gate depth associated with the spectrogram (yellow line). As red represents blood flow toward the probe and blue blood flow away from the probe, both embolic tracks have slopes consistent with their movement direction. Reproduced with permission from Moehring and Spencer.40 HITS indicates high-intensity transient signals; PMD, power M-mode Doppler.

In many instances, it may be unclear whether a HITS represents a true embolus or not; these HITS are referred to as equivocal HITS. HITS need to be differentiated from 2 types of nonembolic signals, namely Doppler speckles and artifacts. Doppler speckles are spontaneous low-intensity fluctuations of the reflected ultrasound and are physiological Doppler flow signals. Speckling is thought to be caused by the nonuniformity of ultrasound fields and corresponds to the movement of RBCs or platelet aggregates that break apart under sheer stress or on contact with the arterial wall.37 Artifacts are associated with movement of the probe, patient movement, and possible electrical interference, and tend to appear as simultaneous bidirectional signals above and below the baseline without the characteristic sound of true emboli. However, emboli can sometimes cause receiver overload resulting in aliasing and bidirectional signals. In these circumstances, HITS representing true emboli cannot be differentiated from artifacts based solely on the consensus committee criteria (duration, unidirectional propagation, intensity, and audible component). The multigate technique to distinguish emboli from artifacts is based on the following principle.38 If a cerebral artery is insonated at 2 (or more) depths along its length, there should be a time delay between the HITS (embolus) seen at the deeper window (more proximal portion of artery) and that seen at the superficial window (more distal portion of artery).38,39 Artifacts, on the other hand, would be expected to be seen simultaneously at both depths. As TCD systems with multiple spectrograms increase the complexity of emboli detection (need to evaluate multiple spectrograms), Moehring and Spencer40 developed power M-mode Doppler (PMD). PMD processes the Doppler data resulting in the simultaneous display of depth from the probe on the vertical axis, time on the horizontal axis, and power of the Doppler shift signal as color intensity at specific depths. Emboli appear as tracks in the PMD display (Figure 3). Rodriguez et al41 have summarized the criteria for differentiating between true HITS, equivocal HITS, Doppler speckles, and artifacts.

Manual Versus Automated Embolus Detection

The gold standard for embolus identification is still the human observer. For research, Doppler monitoring is recorded and evaluated offline by a trained human investigator. This process is very time-consuming, laborious, and mentally strenuous.27 Automated embolus detection software has been developed for commercial systems and laboratory studies to improve diagnostic reliability and reduce the amount of time and effort required for the human observer.42–44 However, the emboli detection software and algorithms of current commercial TCD systems still lack the sensitivity and specificity necessary for clinical use.27,45

Counting Emboli

Quantifying the number of true cerebral emboli per unit time is needed to determine the embolic burden, to identify potential target causes of brain injury, to design future clinical trials intended to reduce embolic load, and to improve patient outcomes. As discussed by Babikian and Wijman,37 commercial TCD systems were primarily developed for measurement of CBFV rather than for emboli detection and quantification. The ability of any given device to detect and quantify emboli depends on the system settings, signal processing algorithms, embolus speed, direction of embolus movement, and position in the cardiac cycle (influences its speed). A critical factor influencing the accuracy of counting emboli is the number and proximity of emboli to each other. HITS events occurring in close succession as clusters and curtains (Figure 1) are difficult to count, so that automated and accurate event separation methods are needed to determine the true number of detectable emboli.42 Even with bilateral MCA monitoring, the true embolic burden will be underestimated because not all cerebral arteries can be monitored simultaneously, emboli pass into side branches of cerebral arteries before reaching the sample window, and limited monitoring duration. The duration of monitoring is also important and may require >30 minutes in some clinical scenarios.46

Differentiation of Gaseous and Solid Emboli

Determination of the true nature of emboli (eg, solid versus gaseous) is needed for clinical decision making.47 The amount of sound reflected at the blood-embolus interface determines the intensity (amplitude) of the returned Doppler signal.48 Since the density of the medium determines acoustic impedance, emboli composed of different materials return signals of different intensities. Air, for example, has an acoustic impedance <1/4000 of blood, resulting in a very large reflection.49 In contrast, the acoustic impedance of solid emboli is closer to that of blood, and this gives a much smaller reflection. In other words, bubbles reflect ultrasound more efficiently than solid materials and therefore result in higher mean intensity increases.48,50,51

Several features have been found to potentially discriminate presumed gaseous from presumed solid emboli: the intensity of the Doppler signal51,52; embolus size (increasing size increases intensity); and sample volume length (SVL) calculated as the product of embolus velocity and duration. SVL is based on the hypothesis that because air reflects the incident ultrasound more strongly, it will therefore be detected over a greater SVL than solid emboli.53 An SVL threshold of 1.28 cm had a sensitivity and specificity of 93% and 97%, respectively, for differentiating between gaseous and particulate emboli. Inadequate dynamic range and poor time resolution of commercially available TCD systems have in part prevented confident discrimination between gaseous and solid emboli.

In the last 2 decades, dual frequency (DF) and frequency modulation (FM) techniques were developed to improve discrimination between gaseous and solid emboli. The frequency dependence of Doppler signal intensity for emboli of different composition was demonstrated by Moehring and Klepper.54 Solid emboli reflect more ultrasound at a higher frequency, whereas gaseous emboli reflect more ultrasound at a lower frequency. The DF technique uses a DF TCD probe (EmboDop, DWL, Singen, Germany) that insonates simultaneously with 2.0 and 2.5 MHz, and automatically classifies embolic signals as gaseous versus solid.42,49 To date, however, DF has not proven reliable when validated using the human observer.49,55 Analysis of FM in the Doppler signals relies on the principle that solid emboli never produce rapid FM.56 When compared head to head in vitro to assess discrimination of large pieces of carotid plaque (0.5–1.55 mm), thrombus mimicking material (0.5–2 mm), and gas bubbles (0.01–2.5 mm), the DF technique yielded a sensitivity of 18% and specificity of 89%, while the FM method had a sensitivity of 72% and specificity of 50%.47 The authors concluded that the DF and FM techniques were unable to confidently distinguish large solid emboli from small gas bubbles.

Another potential method for discriminating between air and solid microemboli is inhalation of 100% O2. Reduction in the inspired nitrogen concentration reduces the partial pressure of nitrogen in the blood, thereby increasing the concentration gradient for diffusion of nitrogen out of a bubble and into the blood. Although O2 will diffuse into the bubble, the higher solubility of O2 versus nitrogen in blood (4.8 times higher) leads to a more rapid reduction in bubble size and dissolution.57 In patients with mechanical heart valves, oxygen-containing microbubbles that are formed through the process of cavitation in areas with turbulent blood flow are short-lived and will rarely be detected in the cerebral circulation in patients breathing a high fraction of inspired oxygen (Fio2).58,59


Cardiac Surgery

One etiology of the diffuse neurologic injury and postoperative cognitive dysfunction (POCD) associated with cardiac surgery, with or without CPB, is cerebral embolism.60 TCD has been used for the detection of cerebral microembolism in cardiac surgery since the late 1980s.61 HITS are frequent during cardiac surgical procedures, particularly with longer duration of CPB.60 In a recent meta-analysis, Indja et al1 found that SBI following cardiac procedures (coronary artery bypass grafting [CABG], off-pump CABG [OPCABG], and aortic valve replacement [AVR]) occur in about 36% of cases and may be a potentially important surrogate marker for brain injury.1 In another meta-analysis of clinical and radiographically-evident brain infarcts, 69.4% of patients undergoing AVR had SBI with only 3.6% having overt clinical strokes.62 In the same analysis, CABG was associated with rates of 27.4% and 2.4% for SBI and overt stroke, respectively.

Another outcome of interest that has been explored is the association between cerebral embolism and POCD, but the data currently remain inconclusive.60,63 Limitations in these studies include small sample sizes and lack of standardized methods of HITS analysis and cognitive outcomes measures. Furthermore, the subclinical nature of any injury caused by microembolism and the multiple etiologies of neurologic injury associated with cardiac surgery makes it difficult to attribute causation.63 Microembolism associated with CPB may be an epiphenomenon for POCD and not represent a target mechanism of brain injury.64 Cardiac surgery and CPB techniques to reduce the risk of systemic embolism include membrane oxygenators, arterial line filters, venous reservoir filters, heparin-bonded circuits, and insufflation of the surgical field with carbon dioxide. TCD is a sensitive technique for the detection of intracardiac right-to-left shunts and has been used in the diagnosis of a patent foramen ovale (PFO).65,66 Detection of a PFO is an important consideration for the risk of cerebral embolism in many cardiac surgical procedures.67

Carotid Endarterectomy

CEA reduces the incidence of stroke in adults with severe (>70%), and to a lesser extent moderate (50%–69%), carotid artery stenosis.68 In the perioperative period, stroke is a known complication due to cerebral thromboembolism from plaque disruption during carotid dissection (80% of infarcts) or cerebral hypoperfusion during carotid cross-clamping (20% of infarcts).69 During CEA, TCD-detected HITS or alterations of CBFV were more common in patients with perioperative strokes compared to those without strokes.70 However, the role of TCD emboli monitoring to guide intraoperative management is not well defined.

TCD has been used in the early postoperative period to identify patients with large numbers of emboli and to administer Dextran-40 to reduce the incidence of stroke.71 However, a Cochrane systematic review found no clear evidence of benefit of hemodilution therapy with respect to mortality or functional outcome for acute ischemic stroke.72

Cardiac Catheterization, Endovascular Stent Placement, and Coiling of Cerebral Aneurysms

Arterial embolism is common during cardiac catheterization and other interventional cardiac procedures.1,62 Interventions that have been associated with HITS production include insertion of intraarterial catheters and guidewires (disruption of atherosclerotic plaque)73,74; left ventriculography and contrast injection into the coronary arteries75; catheter advancement, catheter flushing, contrast injection, and ventriculography during left heart catheterization76; retrograde catheterization of the aortic valve in patients with aortic valve stenosis77,78; transseptal puncture and left disk opening during transcatheter closure of a PFO79; transcatheter aortic valve replacement (TAVR) or implantation (TAVI)80; left atrial ablation of arrhythmias81–83, and following thoracic endovascular aortic repair.84 Whether silent infarcts lead to cognitive or motor impairments in these circumstances remains unclear.

The incidence of stroke with TAVR/I ranges from 3.8% to 4.8% at 30 days.85,86 The risk is significantly greater within the first week and procedural factors are the most important predictor of early stroke.87 Although some studies using TCD found HITS in all patients undergoing TAVR, not all patients develop clinical or SBI.62,77,88 Procedural factors that increase the likelihood of particulate (calcific or atheromatous debris) emboli include excessive catheter manipulation, retrograde catheterization of the aortic valve, manipulation of calcific aortic valves during positioning and implantation of stent valves, longer procedural times, and balloon postdilation (often for paravalvular leaks).87

TCD for detection of microembolic signals (MES) following coiling of unruptured cerebral aneurysms has been used to guide anticoagulation therapy and a significant correlation found between postprocedure MES and ischemic lesions detected by diffusion-weighted imaging.89,90

Mechanical Circulatory Support

Acute brain injuries are a tragic complication for patients on mechanical circulatory support (MCS) with either ECMO or ventricular assist device (VAD). Hypoxia-ischemia, intracranial hemorrhage, and embolic stroke are major mechanisms of acute neurologic injury. Avoiding neurologic injury during MCS is now the key to survival with enhanced long-term neurologic outcomes.91–93

It is well established that cerebral embolism occurs in the setting of MCS.22,94 Cerebral embolism as a complication of MCS can be attributed to thrombosis from low cardiac output and emboli originating from the ECMO circuit, VAD device, or left heart chambers.94–98 Knowledge about the incidence, timing, precipitating factors, and clinical consequences of cerebral emboli in adults on MCS is limited. In a study of 55 adults on venoarterial (VA) or venovenous (VV) ECMO, emboli counts varied with the mode and duration of MCS, with higher HITS rates found during VA-ECMO.22 In 20 adults on left ventricular assist device (LVAD) LVAD support, there was a consistent trend toward enhanced MES activity in patients with high risk of clinical embolism and vice versa.99 To date, adult studies are limited to case reports and small observational studies with significant variability in HITS incidence and detection protocols.100 As with other clinical scenarios, a standardized, rigorous, and more accurate approach to quantifying embolic burden in patients on MCS is needed to further define the nature of potential associations between TCD-detected embolic events and acute brain injuries.


Cardiac Surgery

Few studies have evaluated cerebral embolism in children undergoing cardiac surgery. An early study by O’Brien et al101 used Doppler over the left carotid artery to study 25 children undergoing open repair of congenital heart defects. The median number of detected microemboli was 122 (range, 2–2664) with 42% detected within 3 minutes of release of the aortic cross-clamp. Before CPB, those with right-to-left shunts had significantly more emboli. No relationship between embolism and gross neurologic deficits was found. As part of a multimodality neuromonitoring protocol in infant heart surgery, Naik et al102 evaluated the relationship between the number of embolic signals (HITS) and neurodevelopmental outcome at 3–6 years of age. The median number of HITS with interrogation of the right MCA was 17 (range, 0–55), with no patient having an acute adverse neurologic event. With the small sample size (n = 24) and low embolic count, there was no correlation between HITS and neurodevelopmental outcomes. Rodriguez et al103 randomized children (aged 5 days to 17 years) to Trendelenburg (−15°) or horizontal (0°) head positions during and after standard surgical deairing. HITS were detected in 97% of the 128 patients; the median HITS count was 60 (interquartile range [IQR], 14–189), with neither head position nor surgeon affecting the rate. Electrocardiographic alterations in this study were detected in 27% of patients, and although most resolved spontaneously during heart ejection, 29% persisted. Matte et al104 evaluated gaseous microemboli (GME) in the arterial limb of a CPB circuit with the Emboli Detection and Counting (EDAC) Quantifier (Luna Innovations, Roanoke, VA). With embolic counts indexed to bypass time and body surface area, patients of all sizes had a similar embolic burden. Importantly, air in the venous line increased embolic burden with a 3-fold increase for GME <40 μm and a 5-fold increase for GME >40 μm Another pediatric study with the EDAC found more microemboli in the venous than the arterial line (median, 11,830 vs 1298) and unsurprisingly venous emboli were larger (>40 μm; 59% vs 7%).105 In a small TCD study (n = 4) with EDAC monitors placed on the venous and arterial lines (postfilter), EDAC detected thousands of microemboli in the arterial line with more than 99% being <40 μm in size.106 However, HITS were only detected in 2 patients. Although a relationship between CBFV and neurodevelopmental outcome has been reported, the role of cerebral embolism during pediatric cardiac surgery and its relationship to neurologic outcome warrants further investigation.107

Cardiac Catheterization

HITS detected during cardiac catheterization in children have been associated with clinically evident adverse events.108,109 Like pediatric cardiac surgery, little is known about the prevalence, sequelae, and circumstances of cerebral embolism in children undergoing cardiac catheterization. In 32 children undergoing catheterization for pulmonary or aortic balloon dilations, angiography, or patent ductus arteriosus (PDA) coil placement, HITS were detected in all patients, with a median total HITS count of 44 (95% confidence interval [CI], 27–74).109 Signals were more frequent in patients with septal defects or systemic arterial manipulations, but a small number of HITS were detected in the setting of right-sided procedures and no septal defect. Transcatheter closure of a secundum atrial septal defect (ASD) with the Amplatzer septal occluder (previously AGA Medical, Golden Valley, MN) in children and adults found that MES were significantly higher during left heart catheterization, defect sizing, long sheath placement, and device placement and release.110 The number of MES correlated with the time for device manipulation and placement (r2 = 0.6). In another study, HITS were detected in all patients undergoing ASD closure with the Amplatzer septal occluder; the median HITS count was 67 (range, 9–242) with over 99% temporally associated with catheter manipulation or device placement and release.111 In the studies cited above, the authors described single HITS events. LaRovere et al45 reported much higher HITS counts in a pilot study of 5 children with cardiac disease undergoing cardiac catheterization. With a median monitoring duration of 84 minutes (IQR = 71–109 minutes), the total HITS count for the cohort was 1697. Of these, there were 790 single HITS, 606 HITS within clusters in 3 patients, and 301 HITS within curtains in 4 patients. Events most frequently associated with clusters or curtains were left ventricular angiography (39%), right ventricular angiography (16%), and device placement (16%). In this small study, the authors found that the commercial emboli detection software generated excessive false-positive events. Despite the large number of HITS recorded in these children, none of them developed any clinically evident neurologic injury.

The clinical significance of high HITS counts during pediatric cardiac catherization in the absence of overt injury is unknown. Further investigation of the accuracy of current commercial automated detection and differentiation algorithms to discriminate emboli from artifact is warranted. Although challenging in young children and expensive, magnetic resonance imaging (MRI) and neurocognitive function testing should be part of any study.

Mechanical Circulatory Support

Cerebral infarction and hemorrhage are also significant complications in pediatric patients on MCS with ECMO or VAD. In 49 children who underwent VA ECMO and had a head CT scan, cerebral infarction occurred in 55% and cerebral hemorrhage in 41% of patients.112 The preponderance of infarction was in the MCA territory and frequently bilateral. Infarction patterns on CT were associated with a 7-fold increased risk of seizures on EEG and worse survival (P = .01). HITS have been reported during pediatric VA-ECMO, both spontaneously and in association with bedside interventions (eg, oxygenator change).36 Cerebral embolism may be related to increased ECMO flow rates leading to turbulence and thromboses in the aortic root, within the heart, and at the ECMO cannula tips.113 Depending on the degree of cardiac ejection, flows from the arterial cannula and ventricle can intersect in the aortic arch and result in the propagation of solid emboli.


Strategies to reduce the risk of cerebral embolic injury are lacking, particularly in pediatrics. The adoption of widespread TCD monitoring by clinicians has been hampered by the requirement for a high level of expertise, significant time commitment for monitoring and offline analysis, probe displacement with loss of monitoring, distinguishing artifacts from emboli, precise quantification of embolic counts, differentiation of solid and gaseous emboli, and the relationship to neurologic outcome in asymptomatic patients. Attempts are being made to resolve some of these limitations with improvements in algorithms and autonomous operation with electronic or mechanical steering (as opposed to manual adjustment of the probe).

Novel Algorithm for Accurate Quantification and Differentiation of Emboli

Figure 4.:
TF analysis procedure. A selected embolic HITS (A) is transformed into the TF domain (B). The corresponding TF image is segmented into patches using morphological image processing (C). Individual patches are merged to yield TF subdomains plausibly corresponding to individual embolic segments (D). The final selected subdomains are then transformed back to the time domain, reclassified, and embolic segments are retained (E). Reproduced from Imaduddin et al36 in which this figure appeared Open Access (CC BY 4.0 https://creativecommons.org/licenses/by/4.0/). HITS indicates high-intensity transient signals; TF, time-frequency.

Existing Doppler-based embolus detection and quantification methods, as stated above, have limitations that include inability to differentiate multiple emboli that are simultaneous or close together, excessive false-positive events, and computations over large-signal blocks that restrict real-time monitoring.45,114,115 The authors recently developed a novel single-depth, single-insonation frequency cerebral embolus detection method to address these limitations. Briefly, a weighted-frequency Fourier linear combiner (WFLC) that generates baseline power estimates of received Doppler signals with an adaptively estimated threshold was used to select segments as candidate emboli.36 A time-frequency (TF) segmentation step attempts to separate signatures from emboli that flow into the sample volume concurrently. As shown in Figure 4, a selected embolic HITS is transformed into the TF domain and the TF image is segmented into patches using morphologic image processing. The individual patches are then merged to yield TF subdomains plausibly corresponding to individual embolic segments. Finally, the selected subdomains are transformed back to the time domain, reclassified, and embolic segments are retained. This TF approach led to a median reduction in embolic counts of 64% compared to a commercial 2-depth, DF device (it is important to note that only the signals from the 2 MHz frequency were evaluated). This method is important to both establish accurate embolic counts and for subsequent determination of embolic composition (ie, solid versus gaseous).

Electronic or Mechanical Steering

Operator dependency and probe or patient movement are limitations of conventional TCD systems. The scarcity of emboli in some clinical situations dictates the need for prolonged monitoring, and multiple artifacts in the Doppler spectra produced by probe displacement generates significant difficulties for reliable emboli detection and quantification. The process of reacquiring the vessel and region of interest when the probe is displaced is extremely labor intensive and requires the constant presence of the TCD operator to ensure measurement integrity. Miniaturized and portable TCD systems to address these problems are in varying stages of development.

An electronically steered TCD velocimetry system with machine-learning algorithms for vessel location and tracking have the potential to reduce operator dependency, expedite vessel location, and by updating acoustic focusing of vessels with maximum flow.116 Pietrangelo et al116 have developed a wearable prototype for CBFV measurement that is capable of autonomous vessel search and tracking. Accurate velocimetry of this system has been shown in flow phantom experiments, but human testing is still needed to validate the system.

The TCD-X device (Atys Medical, Soucieu-en-Jarrest, France) is a new, miniaturized, single-gate, and portable device with a robotized probe.114 The orientation of the servo-controlled probe is readjusted automatically when the Doppler signal becomes weak, thereby ensuring the stability of recordings over time. The TCD-X has emboli detection software that provides information (time, intensity, embolus velocity, blood velocity, embolus signal relative duration) for each embolus. The TCD-X device received FDA 510(K) clearance in 2019 (K182175) for ambulatory (outpatient) monitoring of cerebral emboli. The low occurrence of emboli in some patients necessitates monitoring for a prolonged period of time (several hours) rather than the brief monitoring (minutes to an hour or 2) generally performed in the inpatient setting. There is little published clinical data on the evaluation and utility of the TCD-X system. In stroke diagnostics using ultrasound contrast agents, it is not possible to differentiate MES causing the stroke from the ultrasound contrast agent microbubbles. Monitoring with the TCD-X system for 3.5 hours found that sulfur hexafluoride microbubbles (SonoVue, Milan, Italy) persisted for 12–77 minutes (median 40.5 minutes), leading the authors to conclude that MES detection for the identification of stroke-related microemboli should wait at least 77 minutes following the administration of SF6.117 Further evaluation and validation of the TCD-X is still warranted.


TCD sonography is a monitoring modality that allows for the detection of HITS representing cerebral emboli. The lack of a causal pathway in asymptomatic patients does not mean that cerebral microemboli are harmless. Factors associated with the production of macro- and microemboli are both universal (eg, atheroma of aortic arch and carotid arteries) and specific to the clinical setting (eg, CPB). TCD can identify the critical steps associated with MES leading to stroke and SBI, thus helping to develop preventive strategies.83 With respect to microembolism, it is still unknown what combination of embolic load (number of HITS), size, and composition is most injurious. Although SBI are very common and are likely not always “silent,” the use of SBI as a surrogate marker of brain injury has to date been limited by the statistical power of many studies, differences in criteria for differentiating a HITS as an embolus or an artifact, variability with respect to the criteria (including imaging criteria) for SBI and inadequate reported detail of their size, number, and anatomic locations, and differing methodologies for testing cognitive function.1 The ongoing development and evaluation of autonomously steered TCD systems and novel embolus detection algorithms heighten the prospect of operator-independent emboli detection at improved accuracy and facilitating increased use of TCD ultrasonography in a wide range of patients at risk for TIA, SBI, or overt stroke.


Name: Barry D. Kussman, MBBCh, FFA(SA).

Contribution: This author helped with the conception and design of the work, data interpretation, drafted the article, critical revision of the article, and final approval of the version to be published.

Name: Syed M. Imaduddin, SM.

Contribution: This author helped with the data interpretation, drafted the article, revised and reviewed the article, and gave approval for the final version to be published.

Name: Mohammad Hadi Gharedaghi, MD, MPH.

Contribution: This author helped with the data interpretation, drafted the article, critical revision of the article, and final approval of the version to be published.

Name: Thomas Heldt, PhD.

Contribution: This author helped with the data interpretation, drafted the article, revised and reviewed the article, and gave approval for the final version to be published.

Name: Kerri LaRovere, MD.

Contribution: This author helped with the conception and design of the work, data interpretation, drafted the article, revised and reviewed the article, and gave approval for the final version to be published.

This manuscript was handled by: Gregory J. Crosby, MD.


1. Indja B, Woldendorp K, Vallely MP, Grieve SM. Silent brain infarcts following cardiac procedures: a systematic review and meta-analysis. J Am Heart Assoc. 2019;8:e010920.
2. Spencer MP, Campbell SD, Sealey JL, Henry FC, Lindbergh J. Experiments on decompression bubbles in the circulation using ultrasonic and electromagnetic flowmeters. J Occup Med. 1969;11:238–244.
3. Spencer MP, Lawrence GH, Thomas GI, Sauvage LR. The use of ultrasonics in the determination of arterial aeroembolism during open-heart surgery. Ann Thorac Surg. 1969;8:489–497.
4. Muchaidze YA, Syutkina EV. Determination of the linear velocity of the cerebral blood flow in premature infants. Hum Physiol. 1979;5:595–599.
5. Aaslid R, Markwalder TM, Nornes H. Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg. 1982;57:769–774.
6. Bishop CC, Powell S, Rutt D, Browse NL. Transcranial Doppler measurement of middle cerebral artery blood flow velocity: a validation study. Stroke. 1986;17:913–915.
7. Halsey JH, McDowell HA, Gelman S. Transcranial Doppler and rCBF compared in carotid endarterectomy. Stroke. 1986;17:1206–1208.
8. Lindegaard KF, Lundar T, Wiberg J, Sjøberg D, Aaslid R, Nornes H. Variations in middle cerebral artery blood flow investigated with noninvasive transcranial blood velocity measurements. Stroke. 1987;18:1025–1030.
9. Rosenberg AA, Narayanan V, Jones MD Jr. Comparison of anterior cerebral artery blood flow velocity and cerebral blood flow during hypoxia. Pediatr Res. 1985;19:67–70.
10. Taylor GA, Short BL, Walker LK, Traystman RJ. Intracranial blood flow: quantification with duplex Doppler and color Doppler flow US. Radiology. 1990;176:231–236.
11. Trivedi UH, Patel RL, Turtle MR, Venn GE, Chambers DJ. Relative changes in cerebral blood flow during cardiac operations using xenon-133 clearance versus transcranial Doppler sonography. Ann Thorac Surg. 1997;63:167–174.
12. Weyland A, Stephan H, Kazmaier S, et al. Flow velocity measurements as an index of cerebral blood flow. Validity of transcranial Doppler sonographic monitoring during cardiac surgery. Anesthesiology. 1994;81:1401–1410.
13. Alexandrov AV, Sloan MA, Wong LK, et al.; American Society of Neuroimaging Practice Guidelines Committee. Practice standards for transcranial Doppler ultrasound: part I–test performance. J Neuroimaging. 2007;17:11–18.
14. Babikian VL, Wechsler LR. Transcranial Doppler Ultrasonography. 1999.2nd ed. Butterworth-Heinemann;
15. Newell DW, Aaslid R. Transcranial Doppler. 1992.Raven Press;
16. Sharma VK, Wong KS, Alexandrov AV. Transcranial Doppler. Front Neurol Neurosci. 2016;40:124–140.
17. Cobbold RSC. Foundations of Biomedical Ultrasound. 2007.Oxford University Press;
18. Jensen JA. Estimation of Blood Velocities Using Ultrasound: a Signal Processing Approach. 1996.Cambridge University Press;
19. Berger MP, Tegeler CH. Babikian VL, Wechsler LR. , eds. Embolus detection using Doppler ultrasonography. In: Transcranial Doppler Ultrasonography. 1993:Mosby-Year Book; 232–241.
20. Ringelstein EB, Droste DW, Babikian VL, et al. Consensus on microembolus detection by TCD. International Consensus Group on microembolus detection. Stroke. 1998;29:725–729.
21. Kim EH, Lee JH, Song IK. potential role of transfontanelle ultrasound for infants undergoing modified Blalock-Taussig shunt. J Cardiothorac Vasc Anesth. 2018;32:1648–1654.
22. Marinoni M, Migliaccio ML, Trapani S. Cerebral microemboli detected by transcranialDoppler in patients treated with extracorporeal membrane oxygenation. Acta Anaesthesiol Scand. 2016;60:934–944.
23. Kwon JH, Kim JS, Kang DW, Bae KS, Kwon SU. The thickness and texture of temporal bone in brain CT predict acoustic window failure of transcranial Doppler. J Neuroimaging. 2006;16:347–352.
24. Fry FJ, Barger JE. Acoustical properties of the human skull. J Acoust Soc Am. 1978;63:1576–1590.
25. McMahon CJ, McDermott P, Horsfall D, Selvarajah JR, King AT, Vail A. The reproducibility of transcranial Doppler middle cerebral artery velocity measurements: implications for clinical practice. Br J Neurosurg. 2007;21:21–27.
26. Shen Q, Stuart J, Venkatesh B, Wallace J, Lipman J. Inter observer variability of the transcranial Doppler ultrasound technique: impact of lack of practice on the accuracy of measurement. J Clin Monit Comput. 1999;15:179–184.
27. Leunissen T, van Vriesland D, den Ruijter H, Moll F, Mess W, de Borst GJ. Validation of the automated electronic microemboli detection system in patients undergoing carotid endarterectomy. Ultraschall Med. 2018;39:198–205.
28. Markus HS, Ackerstaff R, Babikian V. Intercenter agreement in reading Doppler embolic signals. A multicenter international study. Stroke. 1997;28:1307–1310.
29. Bass A, Krupski WC, Schneider PA, Otis SM, Dilley RB, Bernstein EF. Intraoperative transcranial Doppler: limitations of the method. J Vasc Surg. 1989;10:549–553.
30. Moody DM, Brown WR, Challa VR, Stump DA, Reboussin DM, Legault C. Brain microemboli associated with cardiopulmonary bypass: a histologic and magnetic resonance imaging study. Ann Thorac Surg. 1995;59:1304–1307.
31. Muth CM, Shank ES. Gas embolism. N Engl J Med. 2000;342:476–482.
32. Basic identification criteria of Doppler microembolic signals. Consensus Committee of the Ninth International Cerebral Hemodynamic Symposium. Stroke. 1995;26:1123.
33. Markus HS, Molloy J. Use of a decibel threshold in detecting Doppler embolic signals. Stroke. 1997;28:692–695.
34. Markus HS, Thomson ND, Brown MM. Asymptomatic cerebral embolic signals in symptomatic and asymptomatic carotid artery disease. Brain. 1995;118(pt 4):1005–1011.
35. Babikian VL, Wechsler LR. Transcranial Doppler Ultrasonography. 1993.Mosby-Year Book;
36. Imaduddin SM, LaRovere KL, Kussman BD, Heldt T. A time-frequency approach for cerebral embolic load monitoring. IEEE Trans Biomed Eng. 2020;67:1007–1018.
37. Babikian VL, Wijman CA. Brain embolism monitoring with transcranial Doppler ultrasound. Curr Treat Options Cardiovasc Med. 2003;5:221–232.
38. Georgiadis D, Goeke J, Hill M, et al. A novel technique for identification of Doppler microembolic signals based on the coincidence method: in vitro and in vivo evaluation. Stroke. 1996;27:683–686.
39. 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.
40. Moehring MA, Spencer MP. Power M-mode Doppler (PMD) for observing cerebral blood flow and tracking emboli. Ultrasound Med Biol. 2002;28:49–57.
41. Rodriguez RA, Rubens F, Rodriguez CD, Nathan HJ. Sources of variability in the detection of cerebral emboli with transcranial Doppler during cardiac surgery. J Neuroimaging. 2006;16:126–132.
42. Brucher R, Russell D. Automatic online embolus detection and artifact rejection with the first multifrequency transcranial Doppler. Stroke. 2002;33:1969–1974.
43. Devuyst G, Darbellay GA, Vesin JM. Automatic classification of HITS into artifacts or solid or gaseous emboli by a wavelet representation combined with dual-gate TCD. Stroke. 2001;32:2803–2809.
44. Keunen RW, Hoogenboezem R, Wijnands R, Van den Hengel AC, Ackerstaff RG. Introduction of an embolus detection system based on analysis of the transcranial Doppler audio-signal. J Med Eng Technol. 2008;32:296–304.
45. LaRovere KL, Kapur K, McElhinney DB, Razumovsky A, Kussman BD. Cerebral high-intensity transient signals during pediatric cardiac catheterization: a pilot study using transcranial Doppler ultrasonography. J Neuroimaging. 2017;27:381–387.
46. Droste DW, Decker W, Siemens HJ, Kaps M, Schulte-Altedorneburg G. Variability in occurrence of embolic signals in long term transcranial Doppler recordings. Neurol Res. 1996;18:25–30.
47. Banahan C, Rogerson Z, Rousseau C, Ramnarine KV, Evans DH, Chung EM. Anin vitro comparison of embolus differentiation techniques for clinically significant macroemboli: dual-frequency technique versus frequency modulation method. Ultrasound Med Biol. 2014;40:2642–2654.
48. Markus HS, Brown MM. Differentiation between different pathological cerebral embolic materials using transcranial Doppler in anin vitro model. Stroke. 1993;24:1–5.
49. Russell D, Brucher R. Online automatic discrimination between solid and gaseous cerebral microemboli with the first multifrequency transcranial Doppler. Stroke. 2002;33:1975–1980.
50. Georgiadis D, Hempel A, Baumgartner RW, Zerkowski HR. Doppler microembolic signals during cardiac surgery: comparison between arterial line and middle cerebral artery. J Thorac Cardiovasc Surg. 2003;126:1638–1639.
51. Rodriguez RA, Giachino A, Hosking M, Nathan HJ. Transcranial Doppler characteristics of different embolic materials duringin vivo testing. J Neuroimaging. 2002;12:259–266.
52. Georgiadis D, Mackay TG, Kelman AW, Grosset DG, Wheatley DJ, Lees KR. Differentiation between gaseous and formed embolic materials in vivo. Application in prosthetic heart valve patients. Stroke. 1994;25:1559–1563.
53. Smith JL, Evans DH, Bell PR, Naylor AR. A comparison of four methods for distinguishing Doppler signals from gaseous and particulate emboli. Stroke. 1998;29:1133–1138.
54. Moehring MA, Klepper JR. Pulse Doppler ultrasound detection, characterization and size estimation of emboli in flowing blood. IEEE Trans Biomed Eng. 1994;41:35–44.
55. Markus HS, Punter M. Can transcranial Doppler discriminate between solid and gaseous microemboli? Assessment of a dual-frequency transducer system. Stroke. 2005;36:1731–1734.
56. Smith JL, Evans DH, Naylor AR. Analysis of the frequency modulation present in Doppler ultrasound signals may allow differentiation between particulate and gaseous cerebral emboli. Ultrasound Med Biol. 1997;23:727–734.
57. Kaps M, Hansen J, Weiher M, Tiffert K, Kayser I, Droste DW. Clinically silent microemboli in patients with artificial prosthetic aortic valves are predominantly gaseous and not solid. Stroke. 1997;28:322–325.
58. Droste DW, Hansberg T, Kemény V. Oxygen inhalation can differentiate gaseous from nongaseous microemboli detected by transcranial Doppler ultrasound. Stroke. 1997;28:2453–2456.
59. Thoennissen NH, Allroggen A, Dittrich R. Can Doppler time domain analysis of microembolic signals discriminate between gaseous and solid microemboli in patients with left ventricular assist device? Neurol Res. 2005;27:780–784.
60. Martin KK, Wigginton JB, Babikian VL, Pochay VE, Crittenden MD, Rudolph JL. Intraoperative cerebral high-intensity transient signals and postoperative cognitive function: a systematic review. Am J Surg. 2009;197:55–63.
61. Padayachee TS, Parsons S, Theobold R, Linley J, Gosling RG, Deverall PB. The detection of microemboli in the middle cerebral artery during cardiopulmonary bypass: a transcranial Doppler ultrasound investigation using membrane and bubble oxygenators. Ann Thorac Surg. 1987;44:298–302.
62. Cho SM, Deshpande A, Pasupuleti V, Hernandez AV, Uchino K. Radiographic and clinical brain infarcts in cardiac and diagnostic procedures: a systematic review and meta-analysis. Stroke. 2017;48:2753–2759.
63. Patel N, Minhas JS, Chung EM. Intraoperative embolization and cognitive decline after cardiac surgery: a systematic review. Semin Cardiothorac Vasc Anesth. 2016;20:225–231.
64. Liu YH, Wang DX, Li LH. The effects of cardiopulmonary bypass on the number of cerebral microemboli and the incidence of cognitive dysfunction after coronary artery bypass graft surgery. Anesth Analg. 2009;109:1013–1022.
65. Spencer MP, Moehring MA, Jesurum J, Gray WA, Olsen JV, Reisman M. Power m-mode transcranial Doppler for diagnosis of patent foramen ovale and assessing transcatheter closure. J Neuroimaging. 2004;14:342–349.
66. Teague SM, Sharma MK. Detection of paradoxical cerebral echo contrast embolization by transcranial Doppler ultrasound. Stroke. 1991;22:740–745.
67. Sukernik MR, Mets B, Bennett-Guerrero E. Patent foramen ovale and its significance in the perioperative period. Anesth Analg. 2001;93:1137–1146.
68. Barnett HJ, Taylor DW, Eliasziw M. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med. 1998;339:1415–1425.
69. Krul JM, van Gijn J, Ackerstaff RG, Eikelboom BC, Theodorides T, Vermeulen FE. Site and pathogenesis of infarcts associated with carotid endarterectomy. Stroke. 1989;20:324–328.
70. Udesh R, Natarajan P, Thiagarajan K. Transcranial Doppler monitoring in carotid endarterectomy: a systematic review and meta-analysis. J Ultrasound Med. 2017;36:621–630.
71. Hayes PD, Lloyd AJ, Lennard N. Transcranial Doppler-directed Dextran-40 therapy is a cost-effective method of preventing carotid thrombosis after carotid endarterectomy. Eur J Vasc Endovasc Surg. 2000;19:56–61.
72. Chang TS, Jensen MB. Haemodilution for acute ischaemic stroke. Cochrane Database Syst Rev. 2014:CD000103.
73. Karalis DG, Quinn V, Victor MF, et al. Risk of catheter-related emboli in patients with atherosclerotic debris in the thoracic aorta. Am Heart J. 1996;131:1149–1155.
74. Ramirez G, O’Neill WM Jr, Lambert R, Bloomer HA. Cholesterol embolization: a complication of angiography. Arch Intern Med. 1978;138:1430–1432.
75. Leclercq F, Kassnasrallah S, Cesari JB, et al. Transcranial Doppler detection of cerebral microemboli during left heart catheterization. Cerebrovasc Dis. 2001;12:59–65.
76. Lund C, Nes RB, Ugelstad TP, et al. Cerebral emboli during left heart catheterization may cause acute brain injury. Eur Heart J. 2005;26:1269–1275.
77. Hamon M, Gomes S, Oppenheim C, et al. Cerebral microembolism during cardiac catheterization and risk of acute brain injury: a prospective diffusion-weighted magnetic resonance imaging study. Stroke. 2006;37:2035–2038.
78. Omran H, Schmidt H, Hackenbroch M, et al. Silent and apparent cerebral embolism after retrograde catheterisation of the aortic valve in valvular stenosis: a prospective, randomised study. Lancet. 2003;361:1241–1246.
79. Morandi E, Anzola GP, Casilli F, Onorato E. Silent brain embolism during transcatheter closure of patent foramen ovale: a transcranial Doppler study. Neurol Sci. 2006;27:328–331.
80. Kahlert P, Knipp SC, Schlamann M, et al. Silent and apparent cerebral ischemia after percutaneous transfemoral aortic valve implantation: a diffusion-weighted magnetic resonance imaging study. Circulation. 2010;121:870–878.
81. Gaita F, Caponi D, Pianelli M, et al. Radiofrequency catheter ablation of atrial fibrillation: a cause of silent thromboembolism? Magnetic resonance imaging assessment of cerebral thromboembolism in patients undergoing ablation of atrial fibrillation. Circulation. 2010;122:1667–1673.
82. Miyazaki S, Watanabe T, Kajiyama T, et al. Thromboembolic risks of the procedural process in second-generation cryoballoon ablation procedures: analysis from real-time transcranial Doppler monitoring. Circ Arrhythm Electrophysiol. 2017;10:e005612.
83. Schmidt B, Chun KRJ. More bubble, more trouble? Insights into microembolic events during left atrial ablation. Circ Arrhythm Electrophysiol. 2017;10:e005955.
84. Perera AH, Rudarakanchana N, Monzon L, et al. Cerebral embolization, silent cerebral infarction and neurocognitive decline after thoracic endovascular aortic repair. Br J Surg. 2018;105:366–378.
85. Kapadia S, Agarwal S, Miller DC, et al. Insights into timing, risk factors, and outcomes of stroke and transient ischemic attack after transcatheter aortic valve replacement in the PARTNER trial (placement of aortic transcatheter valves). Circ Cardiovasc Interv. 2016;9:e002981.
86. Kleiman NS, Maini BJ, Reardon MJ, et al. Neurological events following transcatheter aortic valve replacement and their predictors: a report from the CoreValve trials. Circ Cardiovasc Interv. 2016;9:e003551.
87. Dangas GD, Giustino G. Art and science of cerebrovascular event prevention after transcatheter aortic valve replacement. Circ Cardiovasc Interv. 2016;9:e004307.
88. Kahlert P, Al-Rashid F, Döttger P, et al. Cerebral embolization during transcatheter aortic valve implantation: a transcranial Doppler study. Circulation. 2012;126:1245–1255.
89. Cho JH, Kang DH, Kim YW, Park J, Kim YS. Microembolic signal monitoring and the prediction of thromboembolic events following coil embolization of unruptured intracranial aneurysms: diffusion-weighted imaging correlation. Neuroradiology. 2015;57:189–196.
90. Schubert GA, Thomé C, Seiz M, Douville C, Eskridge J. Microembolic signal monitoring after coiling of unruptured cerebral aneurysms: an observational analysis of 123 cases. AJNR Am J Neuroradiol. 2011;32:1386–1391.
91. Barrett CS, Bratton SL, Salvin JW, Laussen PC, Rycus PT, Thiagarajan RR. Neurological injury after extracorporeal membrane oxygenation use to aid pediatric cardiopulmonary resuscitation. Pediatr Crit Care Med. 2009;10:445–451.
92. Schiller RM, Madderom MJ, Reuser JJ, et al. Neuropsychological follow-up after neonatal ECMO. Pediatrics. 2016;138:e20161313.
93. Waitzer E, Riley SP, Perreault T, Shevell MI. Neurologic outcome at school entry for newborns treated with extracorporeal membrane oxygenation for noncardiac indications. J Child Neurol. 2009;24:801–806.
94. Zanatta P, Forti A, Bosco E, et al. Microembolic signals and strategy to prevent gas embolism during extracorporeal membrane oxygenation. J Cardiothorac Surg. 2010;5:5.
95. Asinger RW, Koehler J, Pearce LA, et al. Pathophysiologic correlates of thromboembolism in nonvalvular atrial fibrillation: II. Dense spontaneous echocardiographic contrast (the stroke prevention in atrial fibrillation [SPAF-III] study). J Am Soc Echocardiogr. 1999;12:1088–1096.
96. Dalton HJ, Garcia-Filion P, Holubkov R, et al.; Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network. Association of bleeding and thrombosis with outcome in extracorporeal life support. Pediatr Crit Care Med. 2015;16:167–174.
97. Kwon HM, Park JH, Kim JM, Yoon BW. Mild left ventricular dysfunction is associated with thrombogenicity in cardioembolic stroke. Eur Neurol. 2006;56:217–221.
98. Yang Y, Grosset DG, Li Q, Lees KR. Identification of echocardiographic “smoke” in a bench model with transcranial Doppler ultrasound. Stroke. 2000;31:907–914.
99. Nabavi DG, Stockmann J, Schmid C. Doppler microembolic load predicts risk of thromboembolic complications in Novacor patients. J Thorac Cardiovasc Surg. 2003;126:160–167.
100. Cho SM, Ziai W, Geocadin R, Choi CW, Whitman G. Arterial-sided oxygenator clot and transcranial Doppler ultrasound emboli in venoarterial extracorporeal membrane oxygenation. Ann Thorac Surg. 2019;107:326–327.
101. O’Brien JJ, Butterworth J, Hammon JW, Morris KJ, Phipps JM, Stump DA. Cerebral emboli during cardiac surgery in children. Anesthesiology. 1997;87:1063–1069.
102. Naik RJ, Wagner JB, Chowdhury D. The impact of cerebral embolization during infant cardiac surgery on neurodevelopmental outcomes at intermediate follow-up. Perfusion. 2014;29:443–449.
103. Rodriguez RA, Cornel G, Weerasena NA, Pham B, Splinter WM. Effect of Trendelenburg head position during cardiac deairing on cerebral microemboli in children: a randomized controlled trial. J Thorac Cardiovasc Surg. 2001;121:3–9.
104. Matte GS, Connor KR, Liu H, DiNardo JA, Faraoni D, Pigula F. Arterial limb microemboli during cardiopulmonary bypass:observations from a congenital cardiac surgery practice. J Extra Corpor Technol. 2016;48:5–10.
105. Clark JB, Qiu F, Guan Y, Woitas KR, Myers JL, Undar A. Microemboli detection and classification during pediatric cardiopulmonary bypass. World J Pediatr Congenit Heart Surg. 2011;2:111–114.
106. Wang S, Woitas K, Clark JB, Myers JL, Undar A. Clinical real-time monitoring of gaseous microemboli in pediatric cardiopulmonary bypass. Artif Organs. 2009;33:1026–1030.
107. Cheng HH, Wypij D, Laussen PC. Cerebral blood flow velocity and neurodevelopmental outcome in infants undergoing surgery for congenital heart disease. Ann Thorac Surg. 2014;98:125–132.
108. Esch JJ, Porras D, Bergersen L, Jenkins KJ, Marshall AC. Systemic embolic complications of pulmonary vein angioplasty in children. Pediatr Cardiol. 2015;36:1357–1362.
109. Rodriguez RA, Hosking MC, Duncan WJ, Sinclair B, Teixeira OH, Cornel G. Cerebral blood flow velocities monitored by transcranial Doppler during cardiac catheterizations in children. Cathet Cardiovasc Diagn. 1998;43:282–290.
110. Itoh S, Suda K, Kishimoto S. Microembolic signals measured by transcranial Doppler during transcatheter closure of atrial septal defect using the Amplatzer septal occluder. Cardiol Young. 2011;21:182–186.
111. Wallace S, Døhlen G, Holmstrøm H, Lund C, Russell D. Cerebral microemboli detection and differentiation during transcatheter closure of atrial septal defect in a paediatric population. Cardiol Young. 2015;25:237–244.
112. LaRovere KL, Vonberg FW, Prabhu SP. Patterns of head computed tomography abnormalities during pediatric extracorporeal membrane oxygenation and association with outcomes. Pediatr Neurol. 2017;73:64–70.
113. Janák D, Hála P, Mlcek M, et al. Detection of microembolic signals in the common carotid artery using Doppler sonography in the porcine model of acute heart failure treated by veno-arterial extracorporeal membrane oxygenation. Physiol Res. 2017;66(suppl 4):S529–S536.
114. Guepie BK, Martin M, Lacrosaz V, Almar M, Guibert B, Delachartre P. Sequential emboli detection from ultrasound outpatient data. IEEE J Biomed Health Inform. 2019;23:334–341.
115. Lipperts MG, Sauren LD, Maessen JG, Hoeks AP, Mess WH. Quantification of embolic showers using radio-frequency based TCD analysis. Ultrasound Med Biol. 2009;35:395–402.
116. Pietrangelo SJ, Lee HS, Sodini CG. A wearable transcranial Doppler ultrasound phased array system. Acta Neurochir Suppl. 2018;126:111–114.
117. Aarli SJ, Novotny V, Thomassen L, Kvistad CE, Logallo N, Fromm A. Persistent microembolic signals in the cerebral circulation on transcranial Doppler after intravenous sulfur hexafluoride microbubble infusion. J Neuroimaging. 2020;30:146–149.

Supplemental Digital Content

Copyright © 2021 International Anesthesia Research Society