Figure 4.3 shows the shapes of three consecutive returning pulses as a group of blood cells passes through the sample volume. A short delay is introduced into the signal returning from a given group of cells as they move farther away from the transducer between pulses. An estimate of the velocity can be made by detecting this delay in the echo complex. This shift in the waveform can be understood in two ways. First, it can be considered as a time delay introduced between returning pulses as the group of cells moves. Second, it can be considered as a phase shift of the signal between two pulses scattered from the same sample volume. To understand the concept of a phase shift, consider an example using two clocks. If both the clocks are set to read the same time, both minute hands will be moving at the same speed, giving a frequency of one complete revolution of the clock face an hour. Both clocks will also be in phase with each other (i.e., the minute hand will appear at the 12 on both clocks at the same time), giving a phase shift of zero. If, however, one clock has the minute hand set 30 min behind the other, the minute hands will still complete one revolution of the clock face an hour, but the two clocks will now be out of phase and the clock hands will appear at different places on the clock face. The time delay between the two clocks will be 30 min, or alternatively this delay could be measured as a phase shift, which in this case would be half a cycle of the clock face.

Figure 4.3 Signals from three consecutive pulses returning from a given group of cells as they move farther away from the transducer. Note the short delay introduced.

(Reprinted from Ultrasound in Medicine and Biology, 23/3, Ferrera & DeAngelis, Color Flow Mapping, 1997, with permission from Elsevier)

The delay between the returning ultrasound signals from the first and second pulses shown in Figure 4.3 can be measured in terms of a phase shift. We can see that the second signal is the same shape as the first but is delayed. This is analogous to the clock hands traveling at the same speed but with the second clock being half a cycle behind the first. A similar phase shift can be measured between signals 2 and 3. These phase shifts can be used to estimate the velocity of the blood. This method does not actually measure the Doppler shift frequency; however, the shape of the detected signal is similar to the Doppler shift that would be obtained from a continuous-wave system and therefore can be described by the Doppler equation.

Modern color flow imaging scanners use the phase shift approach, employing a process known as autocorrelation detection to estimate the mean velocity of the flow, relative to the beam. Autocorrelation compares two consecutive pulses returning from a given sample volume to produce an output that is dependent on the phase shift (i.e., dependent on the Doppler frequency). If the echoes are returning from stationary objects there will be no phase shift. The phase shifts between four or more pulses are used to estimate the mean velocity. The more pulses used, the more accurate is the result, as long as the time taken is not so great that the velocity of the blood cells has changed. This relatively small number of pulses required to estimate the mean relative velocity (Doppler shift frequency) enables several color images to be produced every second.

Another method, not currently used by ultrasound manufacturers, uses time domain processing, which employs time delay rather than phase shift to estimate the velocity of blood. From the operator’s perspective, both the frequency and time domain processes produce similar color images.

ELEMENTS OF A COLOR FLOW SCANNER

Figure 4.4 shows the basic elements of a color flow scanner. Before any analysis of the returning echoes is carried out, the signal is filtered by the clutter filter to remove the high-amplitude signals returning from the surrounding stationary tissue and the slow-moving vessel walls, while preserving the low-amplitude signals from the blood. The filtered signal is then analyzed to obtain an estimate of the mean relative velocity (Doppler shift frequency) in each sample volume using the Doppler statistic estimator, as described earlier. Postprocessing is then used to smooth the data in order to produce a less noisy color image. This can be done by combining the data obtained from consecutive images, known as frame-averaging. As each point on the image can only be assigned either a specific color or level of gray, a decision has to be made as to whether to display the pulse echo information or any flow information detected. This involves the process known as blood–tissue discrimination.

Figure 4.4 The basic elements of a color flow scanner.

Blood–tissue discrimination

Generally, the returning pulse echo signals from the vessel lumen are very low in amplitude compared with those from the vessel walls and surrounding tissue. In addition, larger Doppler frequencies are detected from the rapidly moving blood in comparison to the low Doppler frequencies obtained from the slow-moving vessel walls. No Doppler shift would be detected from stationary surrounding tissue. Ultrasound imaging systems are designed with an adjustable control called the ‘color write enable’ or ‘color write priority’ control. This control allows the operator to select the imaging signal intensity above which the gray-scale image is displayed rather than the color information. If a Doppler signal is obtained from an area in which the gray-scale signal is higher than the level set by the operator, the scanner assumes that the Doppler signal results from moving tissue and therefore does not display it. Below this level of gray, providing there is an adequate motion detected, it is assumed that any motion detected originates from blood, and color will overwrite the gray scale in areas where a Doppler signal is detected. If, for instance, the operator wishes to demonstrate flow in a small vessel that does not have an anechoic (echo-free) lumen on the image, the threshold for displaying gray-scale information will need to be increased, thus giving priority to writing color information.

Most systems also have a flash filter. This is designed to remove color flashes, known as flash artifacts, that are generated by rapid movement between the transducer and tissue, such as when the sonographer moves the transducer during scanning.

Color-coding the Doppler information

Having obtained a value of the mean Doppler frequency present in the multiple sample volumes, these data now have to be displayed on the image. This is done by color-coding the Doppler information. The color on the screen has three attributes: luminosity, hue, and saturation. Luminosity is the degree of brightness or shade of the displayed color; hue is the wavelength (i.e., the actual color displayed, from violet through red), and saturation is the degree to which the color is mixed with white light (e.g., from red through light pink, producing up to 20 identifiable tints). These three attributes can be used to produce a variety of color scales, as shown in Figure 4.5, which can be displayed as a bar at the side of the image. The scale usually consists of a different color representing different flow directions, with red often used to show flow toward the transducer and blue depicting flow away from the transducer. Most scanners allow the operator to invert the color scale in order to display flow toward the transducer as blue and flow away as red. This is indicated by inverting the color scale displayed at the side of the image (Fig. 4.5E). It is essential for the operator to be aware of which colors represent which directions of flow within the image, otherwise serious diagnostic errors can occur. Ultrasound scanners provide a range of color scales, and certain scales are more appropriate in particular imaging situations. The various color scales may be selected to accentuate the different parts of the range of detected relative velocities seen in different clinical situations. For example, in an arterial scan, the color scale may accentuate the differences in the upper portion to highlight velocity changes in the higher range of velocities.