The ultimate goal of transcranial embolus detection is to identify potential high-risk embolization with high specificity and validity (Figures 15.3 and 15.4). To achieve this aim several methods were developed to increase the accuracy of emboli detection [16,32,44].

Transcranial color duplex ultrasound investigation can be useful to find the optimal bone window. The suggested frequency is 2–2.5 MHz [35].

Simultaneous monitoring in multiple sample depths (MES appear sequentially in the insonated artery, but artifacts are observed at all depths simultaneously) improves MES/artifact discrimination [33,34] (Tables 15.1 and 15.2).

The recently developed power M-mode Doppler system (PMD) helps to define the optimal depth and the insonation direction for the proper monitoring. PMD constitutes in sum 33 sample gates, two sample volumes per mm, coloring red and blue depending on the flow direction in an M-mode spectrum. The spectrogram displays simultaneous signs of the blood flow velocities of the different vessels. MES appears as high-power sloping tracks on the M-mode spectrum [36,37].

Using a dual-frequency transducer for the analysis of MES results in higher sensitivity and specificity to distinguish between solid and gaseous embolism [8,38].

The validity and reproducibility of the aforementioned methods are better than that of manual techniques.

The patient should be informed, that he/she needs to stay still for the duration of monitoring (Figures 15.5 and 15.6).

The examining practitioner should position him/herself above the patient’s head comfortably, which allows the long-term monitoring. Depending on the doctor’s monitoring technique, the equipment’s display can be placed on the opposite site of the hand holding the transducer, next to the patient’s head.

Using a mobile workstation is helpful.

The monitoring should be performed (if repeated) at the same time of day.

In certain cases, simultaneous electrocardiogram recordings could be recommended (e.g., emboli detection in atrial fibrillation patients).

A preceding transcranial color duplex scan (TCCD) with a probe of 2–2.5 MHz could be helpful for identifying the target vessels and sample volumes’ position.

It is preferable to apply bilateral monitoring, however unilateral monitoring is also possible.

Power M-mode TCD can improve the identification of the MCA (Figure 15.7), and to determine the optimal recording depth, angle and sample volume.

Standardize ultrasound in general (acoustic output, gain, dynamic range, etc.) and probe-dependent (edge enhancement, frame averaging, target frame rate) settings.

To find the temporal bone window place the probe above the zygomatic arch and move it slightly upward or forward [39,40,41].

The headframe should be fixed tightly. If the patient has long hair, it is advisable to wear the hair up.

The MCA usually shows a low-resistance spectrum (similar to the ICA). By decreasing the depth and changing the insonation angle of the transducer the most optimal MCA signal should be detected and monitored. In case of weak signal, increase the gain and the power, and lengthen the sample volume to optimize the quality of ultrasound signal.

Keep the gain as low as possible to eliminate the disturbing background noise.

The transducer should be held in the optimal position to detect the Doppler signs. The headframe should be fixed only thereafter. The quality of headframe is critical, it should prevent the displacement of the probe and decrease the numbers of motion artifacts (Figure 15.8). There are some user-friendly frames:

The optimal depth of the monitoring on the M1 segment of MCA is between 65 and 51 mm.

The registration time should usually last 60 minutes, but, for example, in case of prosthetic valve patients it can be shortened to 30 minutes or even extended in other clinical monitoring situations (e.g., surgical interventions).

Digital recording with validated computer-driven automatic analytical program can help in calculating the types and number of MES [32,36,38,41,42,43,44] (www.spencertechnologies.com/products.html#embolus, www.natus.com/documents/169-436600-1.02%20Emboli%20page_FNL.pdf).

The online analysis of the recorded signals further improves the reliability of monitoring, especially during invasive intervention (cardiac surgery or endarterectomy) [19,45,46].

A robotic probe and detection system (EMS-9UA) was developed for automatic searching and localizing vessel signals, and to automatically track the best position and to restore the lost signal (www.delicasz.com/html/en/).

Uploading of TCD images to a DICOM-compatible PACS is recommended.

Ultrasound device

Transducer type

Insonated artery (e.g., MCA)

Insonation depth (48–58 mm)

Algorithms for signal-intensity measurement

Scale settings (32–100 cm/s)

Detection threshold (≥ 3 dB and ≤9 dB)

Axial extension of sample volume (≥3 and ≤10 mm)

Fast Fourier transformation (FFT) size (number of points used; 64, 128, 256 points)

FFT length (time) (10–100 ms)

FFT transformation overlap (≥50%)

Transmitted ultrasound frequency (2 MHz)

High-pass filter settings (100–200 Hz)

Recording date and time (at least 30 minutes).

Efforts should be made to increase clinical applications of embolus detection (more efficient ultrasound penetration, automatic vessel finding and further optimizing of signal/noise ratio).

Further preclinical and clinical studies should be initiated for improving the reliability of identification of embolus constituents (blood? lipid? air? mixed?).

Further development of user-friendly head- frames that allow long-term embolus monitoring (e.g., surgery) or monitoring during physical exercise.

A new potential direction is the application of an ambulatory TCD system [44] (similarly to the cardiac Holter), which can provide good quality recordings over 5 hours. The emboli recording by TCD on a freely moving person may be helpful for detecting subclinical embolizations provoked by transient events (e.g., paroxysmal atrial fibrillation).

Analyzing the spectra visually in real time is time-consuming, expensive and involves possible errors. The further improvement of already existing automatic detecting systems is eagerly awaited to increase the reliability of the automatic embolic detection systems [14]. The “neuro-fuzzy system” which can achieve a differentiation between emboli and artifacts with a sensitivity, specificity and accuracy of 94.4% is a promising move in this direction [48].

The monitoring of MES can identify responders and nonresponders to therapy. Those patients whose embolic number has not decreased after therapeutic interventions or medical therapy are potentially at high risk for further events. Further prospective studies are needed to answer the question whether embolic detection-guided therapies could help to decrease the risk of stroke or transient ischemic attack. The frequency and number of the repeated monitoring should be determined as well by further clinical trials.

Although the conventional TCD MES detection systems try estimate the number of cerebral microembolisms, the diagnostic accuracy is still not satisfying. Cowe and Evans [41] and Lipperts et al. [11] recommend a new algorithm with high temporal resolution based on radiofrequency signal of a TCD system at least one order of magnitude better than conventional embolic detection methods.

The estimation of the embolus size during emboli detection is still an emerging issue. Banahan et al. elaborated a method to estimate the bubble size based on Doppler embolic size intensity [42]. Under ideal conditions, the in vitro method – in 69% of the cases – could measure correctly the real size of the emboli. Further improvement in this field is needed.

Table 15.3 summarizes the results of tests carried out with embolus detection in different disease groups and methodologies.