Receiving Ultrasound Data

As an ultrasound pulse propagates in the tissue, it interacts with organs and cellular structures. As a result of this interaction, part of the ultrasound energy is reflected back towards the transducer. This is the so-called ultrasound echo. The transducer converts this mechanical energy into an electrical signal.

Shape of the Ultrasound Data as a Function of Time
The transmitted ultrasound pulse typically consists of a short duration (one or a few cycles) sinusoid. As soon as this pulse enters tissue, echos start getting back to the transducer. As the transmitted pulse travels deeper and deeper into tissue, weaker and weaker echos keep getting back to the transducer. Therefore, unlike the transmitted pulse, the echo signal is not of a short duration. As a matter of fact, the echo looks like a continuous sinusoids whose amplitude and phase are changing over time. Because of the similarity of the temporal shape of the echo signal to the temporal shape of Radio Frequency (RF) signals used in telecommunications, the echo signal is called the RF Signal and the received ultrasound data by the transducers, the RF data in the jargon of ultrasound imaging.

In addition to the attenuation of the ultrasound by tissue, which manifests itself as a decreasing amplitude in the RF data, a second major effect is present in the RF data. This is the downshifting of the carrier frequency.

Temporal Shape of the RF data on Sonix Platforms
The temporal shape of the RF signal is determined by a number of different parameters. To compensate for the decreasing amplitude of the RF signal due to tissue attenuation, the (analog) signal is amplified more as the time elapses. This technique is called Time Gain Compensation (TGC) because of the change in the amplifier gain as a function of time. Please see the receive signal path here. In addition to the TGC, there is also a Digital Gain which is applied to the entire RF data at all depths. The way we apply the TGC has evolved with the evolution of our systems. Please refer to Gain Engine for more details.

After the ultrasound pulse has been transmitted, the transducers are electronically switched from transmit to receive. The Sonix will then start recording the RF data and after some time, stop recording them. The start time and stop time for recording the RF data are both adjustable.

The RF data is always sampled at the sampling rate determined by the clock cycle of the system, typically 40, 60, or 80 MHz. However, the data can be downsampled before being returned to the research users. The downsampling is determined by a decimation factor. The data will be downsampled by a factor of 2^decimation factor.

Also see:
 * Gain Engine
 * Texo Parameters for RF Data Shaping

Receive Beam Forming
As with the transmit of ultrasound pulses, a single element transducer usually has a wide field of view. That means when the ultrasound pulse bounces back from different reflectors that are located at the same distance from the transducer, all of the echos arrive at the transducer at the same time. Therefore it would be impossible to distinguish between these reflectors by using the RF data collected by the transducer.

It is ideal that the ultrasound transducer would only see the reflectors that are located on a narrow beam right in front of it. The RF data from such a transducer can be used to generate an image of the tissue features on a narrow line. An array of such transducers would generate a plane image, for instance. As with the transmit of ultrasound pulses, techniques that are used to narrow the field of view of the transducer are called beam forming, or receive beam forming in this case. The two common ways to do so are through physical lenses and electronic beam forming.

One way to narrow the field of view to a smaller area is through the use of physical lenses. Just as an optical lens can be used to focus the optics on a focal zone at a certain distance, an ultrasound lens, attached to the transducer surface can narrow the field of view and focus it on a certain zone. This is called mechanical focusing.

A second way to narrow the field of view is through the use of multiple transducers. This technique is called electronic beam forming. The next figure shows the concept. Five transducer elements are placed side-by-side. For simplicity, let us assume that there is no interaction between the emitted ultrasound pulse and the medium, except from a reflection off the single reflector shown in the figure. As can be seen in this figure, the reflected waves from this reflector arrive at the transducers at different times, because of the difference in the distance from the reflector to the individual transducers. The RF data collected by each transducer shows a pulse at a different time. These are called the pre-beam forming RF data. To focus the received data on the reflector, the pre-beam forming RF data are shifted in time so that the pulses are matched. Then the signals are added. The resulting signal is the beam formed RF data.

As one might notice, there is a caveat in this simplified scenario. we do not know where the reflector is (otherwise we wouldn't be imaging it), so we cannot possibly know how much to shift each of the RF data to get the pulses to match. The key here is to note that we can not possibly be focusing the ultrasound beam on the reflector. The ultrasound beam is out of our hands, and we do not know the location of the reflector. What we can do is to focus the pre-beam forming RF data on a specific point in space.

The pre-beam forming RF data which are also called Channel Data are acquired and stored in some memory by the ultrasound machine. The receive beam forming consists mostly of computational methods to process this data. We first choose a certain point in space (the focal point). Then we calculate the distances hrom the focus point to the individual transducers. These distances are converted to individual time delays for each transducer element by dividing them by the speed of sound (1540 m/s in soft tissue). The pre-beam forming RF data which is residing in the memory is then shifted by its corresponding calculated delay, and then the signals are summed. The resulting sum is the beam formed RF data.

Now if the reflector happens to be at our chosen focal point, it will show up in the beam formed RF data as a big spike. In this case, we say the reflector is in focus. If the reflector is not at our focal point, it will have a blurry appearance on the beam formed RF data.

From this explanation it is clear that we have a choice on where to focus the data. Ideally we would like to focus the data at all the points in the image, one point at a time, to get a clear image at all points. This is one of the main advantages of electronic focusing over mechanical focusing. It is in fact possible, provided that we have: Focusing the ultrasound data at all points throughout the image is called dynamic receive focusing, or dynamic receive beam forming.
 * enough memory to store all the channel data (pre-beam forming RF data)
 * enough bandwidth to transport the data hrom memory to a processing unit and back
 * enough processing capability to do all the beam forming calculations repeatedly for individual points.

In older ultrasound machines, due to hardware limitations, typically a fixed receive focusing was performed. In that scheme, a fixed point located at the center of the elements (center of the aperture) at a selected focal depth was used for the bream forming calculations.

See also: Dynamic focus

Receive Beam Forming on Sonix Platforms
Sonix platforms provide a number of different parameters which can be used for dynamic receive beam-forming using transducer arrays. The SonixRP, SonixTOUCH, and SonixTABLET platforms all perform the receive beam-forming on an FPGA, prior to the transfer of RF-data to the CPU memory of the system. Therefore the pre-beam forming RF data are not available to the user on these systems. Please see the RF signal path here.

SonixDAQ is our platform for collecting pre-beam formed RF data (channel data). It can be used together with our other Sonix platforms, or as a separate module for collecting RF data. However, SonixDAQ does not perform any computation, in particular beam forming, on the RF data.

The Sonix platforms calculate the beam formed RF data by performing a delay and sum algorithm on the pre-beam forming RF data, as described above.

The relative delays for different elements can be determined either automatically, or (manually) by the researcher. In the case of automatic delays, a focal depth needs to be specified for the Sonix software. The Sonix Software then calculates the delays that would focus the RF data at the specified depth.

In the case of manual time delays, the researcher can specify the time delays (needless to say, as a multiple of the clock cycle of the system).

Another adjustable parameter is the aperture, which determines how many elements will be involved in the acquisition of the RF data. These elements will always be chosen evenly on both sides of the center element.

See also: Texo Parameters For Receive Beam Forming

Back to A Reference on Ultrasound Imaging