The history and mechanics of Doppler ultrasonography in the diagnosis of vascular disease are well described by Coleridge-Smith (1996).
This article explains a few simple concepts to guide practitioners in interpretation of the information obtained from a hand-held Doppler probe.
Review of the cardiac cycle
With each cardiac cycle the left ventricle contracts, ejecting blood into the aorta to begin its journey around the body. The blood forced into the aorta meets resistance from the blood already within the vessels and from vasoactivity. This resistance to forward flow is called ‘peripheral resistance’ (Figure 1).
Because of peripheral resistance, not all of the ejected blood can be immediately squeezed down the arterial tree; that which remains distends the walls of the aorta and vessels closest to the heart. The distension of these arteries is made possible by the large amount of elastin within the vessel walls.
This stretching followed by elastic recoil of the arterial walls allows the hydraulic pressure close to the heart to remain at a constant high value, and results in the propagation of a smooth pressure wave.
The pulse that we feel is a pressure wave generated by the heart as it expels blood into the aorta. The character of this pulse is dependent upon many interrelated factors, such as cardiac activity, arterial state and peripheral resistance.
The noises that we hear when using the Doppler machine are generated by the pressure waves, and have been likened to a steam locomotive shunting out of its yard in three-time rhythm or ‘whur–ahh–orr’, denoting a triphasic beat (Figure 2).
The arterial signal is multiphasic, consisting of three parts:
- An initial forward flow rising to a peak at the peak of systole
- A drop from peak systole to reverse flow in diastole
- Finally, another period of forward flow in diastole.
These last two components reflect the elastic recoil of the stored energy resulting from vessel distension at peak systole. This signal is described as triphasic. The typical waveform generated has a mountain range appearance (Figure 2).
Elderly people commonly have a biphasic signal. This characteristically has two parts: the forward flow component at systole, followed by a second forward flow in diastole. This is a ‘normal’ signal; it occurs with ageing and is caused by early atherosclerosis as the vessel walls lose some of their elasticity.
A biphasic signal may also be heard if peripheral resistance is high, e.g. in cold conditions when vasoconstriction is present (Figure 3). Note that there is no reverse flow and the peaks are blunted.
A monophasic signal gives only one sound. This represents a single forward flow component during systole. Monophasic signals may arise when there is vessel stenosis or the vessels are rigid rather than stenosed.
Vessel stenosis: In this situation there is insufficient flow within the vessel during diastole to generate a signal. This denotes vessel disease proximal to the signal site. This may be caused by an isolated stenosis or multiple stenoses.
Other clinical features of vascular disease, such as claudication, low ankle systolic pressure and skin changes may be present (Figure 4). Note that there is no second forward or reverse flow, and a very blunt systolic peak. The sound heard would be lower in pitch than the triphasic or biphasic signal and the signal would be more subtle.
Rigid vessels: In this situation, the sound generated has been likened to that of soldiers marching. The signal is monophasic and the pressures are high. The waveform may be similar to that shown in Figure 5 – in contrast to the smooth rhythmic nature of a normal pulse signal. This is due to the elastic nature of the vessel walls, as demonstrated in the triphasic sound and waveform (Goethals and Brutsaert, 1981).
Pitch: The pitch of the signal is also worthy of note. A high-pitch sound signifies high velocity flow, and a low-pitch sound low velocity flow.
In the diseased state, the signal generated can be almost continuous and sound similar to the venous signal, i.e. ‘a howling gale’ (Figure 6).
Problems commonly encountered in obtaining a good signal are shown in Table 1.
Figures 7-11 show a series of diagrammatic representations of waveforms to illustrate the basic principles of their interpretation. In an article of this nature it is impossible to explore in depth all of the possible variations.
The basic waveform pattern obtained by the Doppler probe is represented as a simple graph, in which the vertical line represents either blood flow velocity or the mean Doppler frequency shift.
The horizontal line re-presents time in seconds (Figure 7). All recordings above the baseline (horizontal line) indicate forward flow and all recordings below the baseline signify reverse flow.
The ‘normal’ arterial waveform pattern shows an initial tall, sharp, mountain peak above the baseline, dropping to a corresponding, but smaller, valley below the baseline, and finally rising to a small hill above the baseline. There then follows a horizontal line running just above baseline before the next mountain peak starts.
The height of the first ‘mountain’, which represents peak systole, is dependent upon cardiac output and peripheral resistance.
In the presence of arterial disease, the reverse flow recording (under the baseline) is lost and the two peaks of forward flow are shallower. When the disease process is more evident, the second forward flow component is lost. Also, in the presence of arterial disease the peak forward flow rises but is notably ‘dampened’, i.e. wider and flatter.
The turbulent flow distal to a stenosis gives rise to a wide range of Doppler shift frequencies, known as spectral broadening. The height of the waveform will be decreased as a consequence of the reduced pressure and velocity within the affected artery.
A waveform that is raised well above the baseline would suggest that the peripheral arteries are dilated (Figure 8).
The presence of a continuous wiggly line below the baseline throughout the recording signifies that the signal is contaminated with venous flow, i.e. the Doppler probe is capturing both arterial and venous blood flow (Figure 11).
Figure 12 shows a series of waveforms illustrating the changes that may be observed at sites before, immediately over, and just after an arterial stenosis. It can be helpful to try to simulate the sounds you would expect to hear as you look at each diagram in turn.
Like all clinical diagnostic modalities, the waveforms recorded are subject to operator error. It is a worthwhile exercise to see what happens when the angle or position of the probe is altered, when pressure is applied, when insufficient coupling medium is used, and when many other variables are introduced. Despite this potential for error, however, the information obtained from a standardised approach can be very useful, particularly in combination with other clinical findings.
It is also useful, when determining the extent of vessel disease, to compare the findings from one limb with those from the other limb and with more proximal signals.
Additional non-invasive diagnostic techniques
Measurement of ankle/ brachial pressure indices (ABPIs) is frequently employed to determine the presence of arterial disease in the clinical situation (Stubbing et al, 1997; Livingston, 1994).
Although ABPI is a very useful tool, it has a major limitation when applied to people with diabetes. Calcification of the tunica media (smooth muscle layer) is common in this patient population (Young et al, 1993). Hence the vessel wall of affected patients is not readily collapsed by the sphygmomanometer cuff, and an erroneously high systolic ankle pressure is frequently obtained.
Diabetes and non-invasive vascular assessments
In patients with diabetes the arteries from the knee to the ankle are aggressively affected by atherosclerosis. In addition, as previously stated, arterial calcification is common and ABPIs are frequently unreliable.
In this situation there are two useful clinical tests that can be performed, which overcome these difficulties and enable assessment of the arterial inflow to the diabetic foot. One is a modified Buerger’s test, called the ‘pole’ test, and the other is toe systolic blood pressure measurement.
This test relies on the fact that pressure within a healthy artery remains relatively constant, irrespective of whether the limb is raised or dependent. In the presence of arterial disease, however, pressure within the limb will drop when the limb is elevated above heart level.
This can be determined in patients by listening to the Doppler signal obtained at a pedal pulse site while gradually moving the limb from a supine position to above heart level. If no change in the signal is detected, there is no significant arterial inflow deficit; however, if the signal fades with a change in phasicity, then arterial disease is indicated.
It is important to recognise that a change in the position of the probe will cause the signal to change, leading to false results.
This test is called the pole test because, traditionally, a graduated pole was placed at the foot end of the patient and the height at which the signal disappeared was recorded. The lower the height the more advanced the arterial disease (Smith et al, 1994).
Systolic pressure measurement is more reliable if performed at the level of the digits, as the digital arteries are less frequently affected by arterial calcification (Holstein, 1984; Orchard and Strandness, 1993).
Many of the techniques for measuring toe pressure are best suited to the research environment rather than to the clinical workplace, e.g. photoplethysmography, strain gauge plethysmography and laser Doppler flowmetry (Neilsen et al, 1972; Lyons et al, 1994; Hoffman et al, 1991).
However, it is possible to detect digital blood flow using a hand-held Doppler with a 10MHz probe, and to obtain a toe pressure using a toe cuff (Carter, 1985). The toe cuff is identical in design to the standard brachial cuff, but much smaller – about 2.5–3.6cm wide (Hoffman et al, 1991; Beinder et al, 1992).
The technique for measuring toe pressure is identical to that used at the ankle, but requires greater dexterity and experience. It is generally easier to measure pressures in the big toe because of its larger size.
It is also important to be aware that toe pressures are lower than those at the ankle. One of the criteria for defining critical limb ischaemia is an ankle pressure of <50mmHg in people without diabetes; in people with diabetes, however, critical limb ischaemia may be defined by a toe pressure of <30mmHg.
The normal value for a toe/brachial index is >0.7. A value of <0.65 would indicate arterial disease (Lyons et al, 1994 ).
Routine vascular assessment in patients with diabetes should include the use of Dopplers, possibly supplemented by waveforms. If arterial calcification is suspected the pole test or toe pressure measurement are invaluable.
The figures are not available full resolution in the online version.