Imaging of all three coronary arteries by transthoracic echocardiography. an illustrated guide

Background Improvements in ultrasound technology has enabled direct, transthoracic visualization of long portions of coronary arteries : the left anterior descending (LAD), circumflex (Cx) and right coronary artery (RCA). Transthoracic measurements of coronary flow velocity were proved to be highly reproducible and correlated with invasive measurements. While clinical applications of transthoracic echocardiography (TTE) of principal coronary arteries are still very limited they will likely grow. The echocardiographers may therefore be interested to know the ultrasonic views, technique of examination and be aware where to look for coronary arteries and how to optimize the images. Methods A step-by-step approach to direct, transthoracic visualization of the LAD, Cx and RCA is presented. The technique of examination is discussed, correlations with basic coronary angiography views and heart anatomy are shown and extensively illustrated with photographs and movie-pictures. Hints concerning optimization of ultrasound images are presented and artifacts of imaging are discussed. Conclusions Direct, transthoracic examination of the LAD, Cx and RCA in adults is possible and may become a useful adjunct to other methods of coronary artery examination but studies are needed to establish its role.


Background
Improvements in ultrasound machines and introduction of the harmonic imaging and the ultrasound contrast agents into clinical practice have enabled direct, transthoracic visualization of long portions of coronary arteries : the left anterior descending (LAD), circumflex (Cx) and right coronary artery (RCA). Most of the studies has dealt with assessment of the middle and distal LAD which can now be visualized in almost every patient [1][2][3][4][5][6][7][8] or the pos-terior interventricular branch, which in skilful hands can be assessed in 3/4 of patients [9]. We have previously shown, that both Cx and RCA can be successfully assessed in a considerable number of sequential adults scheduled for coronary angiography [10,11] and it is known, that they can be studied in children [12,13]. Transthoracic measurements of coronary flow velocity were proved to be highly reproducible and correlated with invasive measurements [1,3,[14][15][16]. While clinical applications of

Methods
We suggest to divide each coronary artery into three segments of approximately the same length (see Figs. 1,2,3,4,5, table 1). The division does not, however, follow exact angiographic classification. The differences are most pronounced with regard to the RCA. As the origin of the circumflex coronary artery cannot always be localized, the left main coronary artery and the proximal part of the left anterior descending artery form a common segment.
A modern, high quality ultrasound machine is required for successful scanning. In most transthoracic, echocardiographic examinations (TTE) performed in our laboratory we have used an Acuson Sequoia 512 ultrasound instrument (Acuson Corp. Mountainview, Ca). We are using either 5.0 or 3.5 MHz narrow-band sector transducers in second harmonic mode for B-mode examination while both color Doppler coronary artery mapping and spectral Doppler coronary flow velocity assessment are performed at 2.5 or 2 MHz. The coronary arteries can, however, be imaged with ultrasound instruments made by other manufacturers and with both broadband and high frequency transducers.
We believe that the harmonic imaging considerably facilitates the examination. Many investigators [1,3,4,17] but not all [11] concluded, that ultrasound contrast agents provided a substantial improvement in the quality of Doppler coronary artery images. While technology of An artists drawing of the heart, right antero-oblique (RAO) view ultrasound contrast agents is still developing and experience with their most effective use is growing the exact role of ultrasound contrast agents in coronary artery examination remains to be established.

Visualisation of the left anterior descending coronary artery
With the patient in left decubitus position the transducer is placed at the left parasternal location and a parasternal short axis view of the great vessels is obtained. Slight change of the imaging plane so that it traverses the heart immediately below the pulmonary trunk allows for visualization of the left main (LM) and proximal LAD. These arteries may be relatively easily visualized in B-mode while Doppler examination is usually required to locate the more distal parts of the LAD (Fig. 6,7,8,9, movie 1, 2 [see Additional file 1, 2]). Depending on anatomical relations in a particular patient, sometimes it is easier to assess the proximal part of the LAD in a modified parasternal long axis views. The artery may be seen there as it traverses below the pulmonary trunk towards the interventricular groove (Fig. 10). The LM and proximal LAD may also be viewed using a modified apical five chamber view (traversing anterior wall of the heart) within the area lateral to the sinus of Valsalva (Fig. 11,12).
The middle and distal LAD may be assessed by placing the transducer in low parasternal position. The interventricular groove is located in short axis view and then the transducer is either tilted towards the base of the heart or rotated to obtain a modified long axis view (aligned parallel to the groove). Long axis view of the LAD can be obtained using either of these views (Fig. 13,14,15,16, movie 3 [see Additional file 3]). The middle and distal LAD can also be visualized using a modified apical 5chamber view (traversing anterior wall of the heart). The entire LAD together with some diagonal branches can sometimes be assessed in this view (Fig. 12). The apical window is often used for examination of distal LAD at the apex (Fig. 17,18) and even on the inferior wall of the ventricle (Fig. 19,20). Apical window is usually the best to obtain low Doppler angles for velocity measurements in the LAD. The LM and the full length of the LAD may also be visualized in apical 3-chamber view, although the motion artifacts, proximity of the lung and winding of the LAD along its way down to the apex make examination difficult.
The patient may then be placed in supine position. The transducer is moved to the subcostal location and a long axis view of the heart is obtained. Then the transducer is tilted cephalad aiming at the anterior wall of the left ventricle where long segments of the LAD can be visualized. However, the quality of LAD images obtained in this view is usually low and Doppler angle is suboptimal (Fig. 21,22).

Visualisation of the left circumflex coronary artery
Since it may be assumed, that the circumflex coronary artery leaves the coronary sulcus at its lowermost point (as can be seen on parasternal short axis view), the left part of the sulcus between aorta and lowermost point of the sulcus may be divided into two sections of approximately the same length, and segments of the circumflex artery within respective sections of the sulcus may be considered proximal and middle Cx (Fig. 3,4).
The Cx is visualized in parasternal short axis view with imaging plane traversing the heart immediately below the pulmonary trunk (Fig. 23, movie 1, 2, 4 [see Additional file 1, 2, 4]). Sometimes it is better seen when the probe is directed just below the left atrial appendage (Fig. 24). Usually a slight change of probe orientation is necessary to visualize the middle Cx which runs down the coronary sulcus at the border between the atrium and the ventricle ( The transducer may then be moved to the apex and a modified five chamber view (traversing anterior wall of the heart) is obtained. Area lateral to the sinus of Valsalva is carefully searched and one may find there the Cx originating from the LM. A reliable measurement of the flow velocity within the initial part of the Cx may then be per-formed. Further segment of the Cx can also be visualized in B-mode, but as this part of Cx with this ultrasonic window runs almost perpendicular to the ultrasonic beam and is located several centimeters away from the transducer, this particular view is of limited use for the assessment of the Cx (Fig. 28).
We have never managed to visualize the distal Cx -at the infero-lateral wall of the ventricle.

Visualisation of the right coronary artery
The parasternal short axis plane may be used to visualize the proximal RCA (Fig. 29). It may be relatively easily visualized in B-mode (Fig. 30) while Doppler examination is Parasternal short axis view at the level of mitral anulus Figure 4 Parasternal short axis view at the level of mitral anulus. It is usually not possible to visualize proximal and middle part of the Cx in a single cross sectional view due to different spatial orientation of proximal and middle segments of the artery. Here, approximate location of proximal segment and location of middle segment in relation to the cross section of left ventricle at the level of mitral anulus is shown.
usually required to locate its middle and distal parts. Depending on anatomical relations in a particular patient, sometimes it is easier to assess the proximal RCA in a modified parasternal long axis views. The proximal RCA may be seen in this view running along the anterior wall of the aorta (Fig. 31). The proximal RCA can infrequently be visualized using apical window, but this view allows for recording of the flow velocity spectrum (Fig. 32, 33).
The middle RCA is visualized with the patient in supine position. The transducer is placed in subcostal location and the short axis view of the heart is obtained. The anterior wall of right ventricle at the coronary sulcus is carefully scanned and it is where good quality middle RCA images at low Doppler angles can be obtained ( The distal RCA can be assessed using apical window. An apical four chamber view is obtained and then the transducer is tilted more perpendicular to the chest so that the imaging plane traverses the inferior wall of the heart and the posterior interventricular groove. Good quality images of the right posterior descending coronary arteryusually distal part of the RCA -can be obtained in this way. Alternatively, a modified 2-chamber view (crossing the heart along the posterior interventricular groove) can be used (Fig. 37,38,39, movie 8 [see Additional file 8]).
Angiography of the right coronary artery (RCA) in LAO projection Figure 5 Angiography of the right coronary artery (RCA) in LAO projection. Division of the RCA into segments: proximal (P), middle (M) and distal (D) -the posterior descending branch  Proximal parts of the LAD and Cx in color Doppler; modi-fied parasternal long axis view Figure 10 Proximal parts of the LAD and Cx in color Doppler; modified parasternal long axis view to a significant LAD stenosis. Flow velocities vary considerably depending on many factors: the heart rate, blood pressure, thickness of myocardium, coexistent heart disease, medications, perhaps even on vessel involvement in the formation of collateral circulation [19][20][21]. It increases at the site of stenosis. In a prospective study of sequential patients referred for coronary angiography we found maximal flow velocity within the LAD, Cx and RCA in the range of 0,7 m/s for proximal parts of non-stenosed arteries and somewhat lesser values for their distal parts [10], but there are wide variations in the literature regarding maximal flow velocity in non-stenosed coronary arteries.
The spectrum recorded on transthoracic coronary artery examination is frequently devoid of a discernible systolic component, most probably due to movements of the heart (Fig. 41). For this reason the estimation of mean velocity and velocity time integral by TTE examination of coronary arteries may be impossible.

Hints concerning machine settings
There is a choice of ultrasound transducers available which may be used to examine the coronary arteries. The higher frequency transducers provide high resolution Bmode images and improved sensitivity of Doppler examination (both color and pulsed wave [PW] spectral Doppler). The high frequency transducers may be most successfully used when apical area is scanned -i.e. for examination of the distal LAD. The lower frequency transducers provide a good penetration and higher Nyquist limit for PW-spectral and color Doppler examination at the expense of lower resolution and lower Left main coronary artery (LCA) arising from aorta (AO) in B-mode; modified apical 5-chamber view Figure 11 Left main coronary artery (LCA) arising from aorta (AO) in B-mode; modified apical 5-chamber view.
sensitivity of Doppler exam. Nevertheless, they seem to be the optimal choice for scanning of RCA, Cx and proximal and middle LAD.
We suggest optimal quality B-mode image should be obtained before one starts to look for coronary arteries. When poor quality of B-mode image does not allow for visualization of coronary arteries one may decrease the dynamic range of B-mode scanning, change the ultrasonic window or use the harmonic mode for B-mode, which we believe greatly facilitates scanning of the coronary arteries.
Then the color Doppler is turned on and the Nyquist limit for color examination is in most instances adjusted to 0,4 -0,5 m/s. Too little the Nyquist limit the more color Doppler artifacts obscuring the images are encountered. On the other hand, too high a Nyquist will make detection of low velocity flow within the coronary arteries impossible. Gain is increased until large amount of color artifacts almost obliterate the image. It is then turned down a little and the area where the artery should be located is carefully scanned. Further gain and Nyquist limit adjustments are performed when flow signal within the artery is detected. The size of the area where color Doppler examination is carried out should be kept at minimum. Some ultrasound machine manufacturers suggest a special color map should be used for coronary artery examination -a map, which does not discriminate between velocities to and away from the transducer. We, however, believe a regular color map is perfectly sufficient to this aim. If the artery cannot be localized, change of ultrasonic window (change of the distance between the vessel and transducer, change of Doppler angle), further optimization of B-mode image or readjustment of color settings (gain, size of the area, where color examination is carried out or Nyquist limit) or change of transducer frequency should be considered. If these fail an ultrasound contrast agent may be tried.
Usually a small part of the artery can be seen at first. Then, by moving step-by-step up and down the course of the vessel the entire artery (or its long part) may be assessed (Fig. 42, 43, movie 9 [see Additional file 9]).
Once the vessel is located velocity spectrum is recorded.
As there is no general agreement on which Doppler angle is optimal for PW spectral Doppler (duplex) examination it is universally accepted that for the duplex scanning the angle should never exceed 60 degrees. The lower the angle the less the possible error introduced by any inaccuracies of Doppler angle estimation but at the possible price of sensitivity of recording of higher velocities. The latter may be particularly important in very tight stenoses where Doppler signal intensity may not be strong enough to allow for correct measurement of maximal velocity. In such cases it may be advisable to change the window so that the Doppler angle increases towards (but does not exceed) 60 degrees.
We usually use a 5 mm gate for pulsed Doppler (to narrow a gate would decrease the signal intensity and too wide would make the signal more "noisy"). Smaller gate (3 mm) may be used when distal segment of the LAD is examined. We generally use the manufacturer default Doppler filter setting (as for regular echocardiographic examination).

Artifacts of scanning. Sources of diagnostic errors
Artifacts pose an important problem which make the examination exceptionally difficult, especially when the Cx or RCA are scanned. Most of them are caused by movements of the heart.
Often the flow of physiological amount of fluid within the pericardial sac generates a strong signal which may be confused with flow within a coronary artery (Fig.  43,44,45,46,47, movie 10 [see Additional file 10]). The flow within the pericardial sac is most pronounced in systole, while the coronary artery flow is predominantly diastolic and this should allow for correct interpretation of images.
Left coronary artery in color Doppler; modified apical 5-chamber view Figure 12 Left coronary artery in color Doppler; modified apical 5chamber view. Parts of the left main and LAD coronary arteries are seen. Color aliasing indicating local stenosis of proximal (LAD PROX STEN) and middle LAD (LAD MID STEN) are seen. Color aliasing was also noted in proximal part of a diagonal branch (not shown in this frame). Duplex criteria of significant stenosis were met for all stenoses, the diagnosis was subsequently confirmed by coronary angiography We have on a few occasions observed a characteristic artifact mimicking a very high velocity flow within the middle or distal LAD, which can be recorded both in color and spectral Doppler (Fig. 48,49, movie 11 [see Additional file 11]). When carefully examining the velocity spectrum one may notice then, that: 1. the signal is not confined to diastole -it is rather recorded in a random way, without any definite relation to the cardiac cycle and 2. the maximal velocity cannot be recorded due to aliasing.
When middle LAD is scanned some extracardiac vessels as well as a cardiac vein can be visualized on color Doppler examination, which may be confused with the LAD. However, spectral Doppler examination allows for proper distinction: the extracardiac arteries show a spectrum typical for peripheral arteries with predominantly systolic flow while veins show a steady, systolic-diastolic, low velocity flow which would cessate or significantly decrease on Valsalva maneuver. The left internal thoracic artery (LIMA) Figure 13 Angiography of the left coronary artery in RAO projection with caudal angulation. Middle segment of the LAD is marked by red circle for correlation with echocardiographic image may be confused with the LAD in cases, in which the LIMA was used for a coronary artery by-pass. The velocity spectrum within the LIMA has then a prominent diastolic component typical for the coronary arteries which may be misleading for a beginner. However, the systolic part of velocity spectrum recorded in LIMA is typical for a peripheral artery. Although the systolic component successively diminishes between the origin of the artery internal mammary artery and its anastomosis with the coronary artery, it is more pronounced, than the diastolic component and therefore easily discernible.

Angiography of the left coronary artery in RAO projection with caudal angulation
The posterior descending coronary artery which usually is a distal branch of the right coronary artery can arise from the Cx when left coronary artery is dominant, which may cause diagnostic mistakes.
Some large branches of coronary arteries -the intermediate coronary artery (Fig. 8, movie 1 [see Additional file 1]), diagonal and marginal branches can sometimes be visualized and confused with principal coronary arteries, most notably the LAD and Cx. It is especially likely when the principal artery is occluded and a branch enlarges and takes over its function. We have occasionally encountered this problem with regard to proximal LAD and proximal Cx. It is easier to properly locate the LAD at its middle part, as it is clearly localized in the anterior interventricular groove. By moving step-by-step towards aorta or downwards to the apex without loosing the LAD from sight one may correctly identify the proximal and distal LAD. The fact, that distal LAD is the only artery which is visualized from a strictly apical transducer posi- tion and that it "winds" to the inferior wall of the heart is another clue (Fig. 20).

Middle LAD in color Doppler
The differentiation between various vessels seen at the lateral and inferior wall of the ventricle may be challenging. Figure 15 Angiography of the left coronary artery in RAO projection with caudal angulation. Distal segment of the LAD at the anterior wall of the ventricle is marked by red circle for correlation with echocardiographic image Improper machine settings, most notably choosing to high or too low Nuquist limit for color Doppler may result in failure to visualize the coronary arteries.

Discussion: Practical aspects of TTE examination of coronary arteries
Color duplex ultrasound examination is an invaluable tool for assessment of peripheral arteries. While studies are needed to determine if the TTE examination of long segments of coronary arteries will become a useful ancillary test in cardiology we can summarize our 5-year experience during which we have performed some 800 scans of the LAD, the Cx and /or the RCA as follows [10,11] : 1. The examination is technically demanding, but the feasibility of visualization of the LAD is high and moderate with regard to the Cx and RCA [ [9][10][11]22], table 2].
2. Although there are no firmly established Doppler criteria of coronary artery stenosis we have been quite successful using the criteria valid for peripheral arteries i.e. when at least doubling of maximal flow velocity within stenosis with regard to maximal flow velocity immediately proximal to stenosis was taken as a sign of >50% reduction of vessel diameter. Similar results were obtained by Hozumi et al. who used a ratio of mean velocities of 2.22 combined with local color aliasing as a sign of restenosis after percutaneous coronary intervention within the LAD [22]. Retrospective analysis of our results has shown, that a local maximal flow velocity > 1.5 m/s is also a reasonably Distal LAD at the anterior wall of the left ventricle in color Doppler Figure 16 Distal LAD at the anterior wall of the left ventricle in color Doppler. Modified low parasternal long axis view accurate sign of stenosis. The velocity ratio criterion seems more reliable than the criterion of maximal flow velocity, but the former may not be used in ostial lesions as there is no reference flow velocity for comparisons.
3. Once the artery can be visualized and spectral Doppler examination can be performed the accuracy of assessment (stenosis present / absent) is high. While our results await independent confirmation, we have shown that direct, Angiography of the left coronary artery in RAO projection with caudal angulation Figure 17 Angiography of the left coronary artery in RAO projection with caudal angulation. Distal segment of the LAD at the apex is marked by red circle for correlation with echocardiographic image transthoracic color Doppler examination of coronary arteries can be used to monitor restenosis after PCI within the LAD, Cx or RCA (Table 3). 4. It helps to know the anatomical details of coronary arteries (i.e. to review the coronary angiography) before the TTE of coronary arteries is performed. It can be easily done if TTE is used to monitor restenosis as films performed during the PCI are readily available. Our present results [11] are better than our previous [10], which may be explained by a learning curve, but also by the availability of coronary angiograms to the echocardiographer at the time of scanning. Most diagnostic errors relate to improper identification of principal coronary arteriesi.e. taking a strong branch -septal or diagonal for the LAD and marginal for the Cx.

5.
Occasionally the diagnosis of coronary artery stenosis is so easy and straightforward that it can be made during a regular echocardiographic examination, taking only a few minutes to complete and not causing any meaningful prolongation of scanning (Fig. 50,51,52

Authors' contributions
MK and WB carried out the echocardiographic examinations prepared the pictorial materials and drafted the manuscript.
PPD participated in some of the studies and reviewed the manuscript.

All authors read and approved the final manuscript
Distal LAD at the apex in color Doppler using a 7 MHz transducer Figure 18 Distal LAD at the apex in color Doppler using a 7 MHz transducer. Apical 3-chamber view. Courtesy of Acuson Corp. Angiography of the left coronary artery in RAO projection with caudal angulation Figure 19 Angiography of the left coronary artery in RAO projection with caudal angulation. Distal segment of the LAD at the inferior wall of the heart is marked by red circle for correlation with echocardiographic image Distal LAD at the inferior wall of the heart in color Doppler Figure 20 Distal LAD at the inferior wall of the heart in color Doppler. Apical 3-chamber view. Blood flow is depicted in blue.