Read untitled text version

Ultrasound Physics, Artifacts, and Equipment

Laura E. Ziegler, DVM, Dip. ACVR

Why physics?

· Helps to understand WHY things look the way they do · Provides a better understanding of the limitations of ultrasound · Is vital to understanding artifacts (and potentially how to correct them) · Better able to understand "sales talk" and make rational decisions

What is ultrasound?

· Ultrasound is defined as a sound wave with a frequency >20 kHz · Diagnostic ultrasound transducers generally are in the range of 1-20 MHz · Frequency, wavelength, and velocity are related by the wave equation: =f x


How is the sound generated?

· Piezoelectric crystals in the transducer face vibrate at a specific range of frequencies when an electric current is put through them · This vibration results in alternating compression and rarefacation of the surrounding material, which is then transmitted into the patient

How is the sound generated?

What happens to the sound in the body?

· Sound passing through a material can do one of four things:

­ ­ ­ ­ Be absorbed Be transmitted Reflect Refract


Transmission and reflection

· Velocity of transmission varies slightly from tissue to tissue, but the average is 1540 m/s · Acoustic impedance refers to the ability of a material to transmit sound

­ Air 400 kg/m2-sec ­ Liver 1.65 X 106 kg/m2-sec ­ Bone 7.8 X 106 kg/m2-sec

Transmission and reflection

· If there is a DIFFERENCE in impedance at an interface, some of the sound will be reflected · The bigger the difference, the bigger the reflection

Transmission and reflection

· The "rules" of reflection are the same as for light, an incident wave hitting a smooth angled surface will NOT be reflected back to the source · However, in medical ultrasound, we are very rarely dealing with truly smooth surfaces


Absorption and scatter

· Scatter is reflection or refraction that results in the sound wave never coming back · Absorption takes place when the energy of the ultrasound wave is transformed into heat · The combination of absorption and scatter is referred to as Attenuation (decibels/cm)

· Amount of attentuation is dependent on the material (e.g. fluid vs. liver vs. bone), and the transducer frequency


Attenuation pic

Absorption and scatter

· Degree of attenuation in soft tissue has a nearly linear relationship with frequency · Higher the frequency, higher the attenuation · Higher the attenuation, less penetration



· Discussed in more detail when we talk about artifacts · Basically, when there is a angled interface between two materials, the sound wave gets bent

Interaction of sound with matter

Then what?

· Attenuated sound waves are never seen again · Transmitted sound waves continue on to other opportunities to be reflected, refracted, or transmitted · Sound waves that are reflected (more or less) straight back to the transducer will vibrate the crystals, which generates an electrical signal


Then what?

· The computer then makes a calculation based on the "known" velocity of sound in tissue

­ If the echo took 0.0325 msec to return,

· 154000cm/s x 0.0325 m sec=5.005 cm

· It will plot the whiteness of the "spot" in proportion to the strength of the signal (measured in decibels)

­ 2.0025 cm deep in the above example

Then what?

· The computer also assumes a certain amount of attenuation takes place, so will "boost" signals that have taken longer to come back (time-gain compensation or TGC) · Deeper echoes will be depicted as brighter than more shallow echoes of the same absolute strength


· The transducer is in "sending" mode only about 1/1000th of the time it is on. The remaining 999/1000th s are spent "listening" · If it didn't, the computer would have no way of knowing which echoes came from which outgoing pulses and couldn't do it's math · Exception is continuous-wave Doppler

Pulse Repetition Frequency

· aka PRF · Refers to how frequently a given pulse is repeated

· Easiest way to think of PRF is to think of a strobe light--the faster the blink, the more normal everything looks

Types of ultrasound

· A-mode--first type developed, a.k.a. Amplitude mode

­ Is a stationary "ice pick" view ­ Reflectors are depicted as spikes whose height reflects the amplitude of the return signal ­ Sometimes used in ophtho. ­ Primitive A-mode devices can be used for pregnancy checks in livestock


Types of ultrasound

· B-mode--so called because it came after Amode (supposedly stands for "brightness")

­ What we're most familiar with is "real-time" Bmode ­ Reflectors are depicted as dots with a brightness corresponding to the amplitude ­ The 2-dimensional image is formed by numerous static "ice picks" displayed at once



Types of ultrasound

· M-mode--motion mode

­ Consists of numerous "ice pick" views taken over time ­ Allows for very good temporal resolution as the PRF can be very fast ­ Commonly used in cardiology


Types of ultrasound

· Doppler--detects motion of reflectors in the field

­ When sound strikes a moving object a Doppler shift occurs ­ Frequency shift =2f(/c)cos ­ If the sound wave is perpendicular to the moving object, no Doppler shift occurs ­ Interestingly, the shift is typically in the range of human hearing


Types of ultrasound

· Continuous wave Doppler

­ One crystal constantly sends, another constantly receives ­ Allows evaluation of very high velocities (good temporal resolution) because the PRF is essentially infinite ­ Use is essentially limited to cardiology and anesthesia monitoring

Types of ultrasound

· Pulse wave Doppler

­ Doppler information is generated from a small "gate" that is interrogated ­ Information is displayed graphically and allows for simple measurement, also audible ­ Usually a "duplex" display is used


Pulse Doppler

Types of ultrasound

· Color Doppler

­ A fairly large "gate" is analyzed and the Doppler shift is depicted as shades of color ­ Traditionally, shades of red are reflectors moving toward the transducer, "blue away" ­ Power Doppler is somewhat more sensitive, but generally throws out directional information

Color Doppler


What makes a good image?

· Axial resolution is the ability to discriminate between two objects along the axis of the beam · It is dependent on pulse length (which is dependent on the frequency and the pulse duration) · Higher frequency=shorter wavelength=shorter pulse length=better resolution · Shorter pulse duration=better resolution


What makes a good image?

· Lateral resolution is dependent on the beam width · Lateral resolution will be better at the focal zone · Digitally focused transducers are more flexible than non-focused or mechanical transducers


What makes a good image?

· Temporal resolution (particularly important in high-motion exams such as echocardiography) is the ability to resolve events in time · Temporal resolution is dependent on pulse repetition frequency (PRF)--the length of time it takes for one pulse to be repeated

· If you use narrower "ice picks" lateral resolution will improve, but you will need more of them to fill the screen (poorer temporal resolution) · If you target a smaller field of view, you won't need as many "ice picks" to fill it (better PRF) without decreasing your lateral resolution

· Multiple focal zones can improve lateral resolution, but each zone is a devisor of PRF (i.e. three focal zones takes three times as long) · Keep in mind that if the object you are examining is not moving, you can have as bad a PRF as you like


What makes a good image?

· Display

­ Monitor resolution ­ Grey scale

· Grey scale "quantity" can't be adjusted, but "quality" might be

­ Brightness/contrast ­ Time-gain compensation (TGC)

· TGC partially adjustable by user


· Echogenicity is ALWAYS described in relation to something else · Sometimes the something else is assumed · Anechoic means NO echoes


· Some artifacts are diagnostically useful · However, most artifacts are distracting, and may even be clinically treacherous · Origin of all artifacts is based on physics, computer stupidity, or some combination



· Shadowing

­ Created by either complete absorption or complete reflection of the incident sound wave ­ Generally indicative of gas, mineral, metal, wood, asparagus, etc. ­ Large objects, or smaller ones with smooth sides, or large accumulations of small objects will have a "clean" shadow ­ Smaller, irregular or less dense objects or substances will have a "dirty" shadow




· Distal enhancement (a.k.a. through transmission)

­ Sound passes through material that is less attenuating than the computer assumes ­ The computer boosts the signal the same that it would normally (TGC) ­ Result is a "stripe" of increased echogenicity beyond the object ­ Generally BUT NOT ALWAYS indicates a fluid-filled structure (e.g. a cyst)

Distal enhancement


· Reverberation (a.k.a. "ringdown", "bang", "comet tail")

­ Two mechanisms of creation ­ Reflection hits transducer and reflects off of the transducer, back to object, back to transducer and so forth ­ Object present in body "rings" when struck by sound




· Reverberation

­ Each portion of the reflection or "ringing" that is detected by the transducer is depicted as a bright line ­ The bright lines show up at regularly spaced intervals corresponding to either the distance of the round trip or the frequency of the "ringing"






· Refraction artifact, also called edge shadowing, occurs distal to a curvilinear interface · Recall from physics: when a wave strikes an interface at an angle with two different propagation speeds, it will be deflected toward the "slower" substrate



· Refraction is most commonly a simple annoyance · Using multiple scan planes will enable examination of objects in the "shadow" · Occasionally, will trap the unwary into thinking there is a defect in the urinary bladder or diaphragm (generally only if there is fluid on either side)



Mirror image

· Occurs when there is a strong reflective interface

­ Sound waves bounce off the "mirror", start to return, then bounce off a lesser reflector ­ These waves return to the "mirror", then finally reflect back to the transducer

· Computer only knows that the echo has taken "x" longer to get back, so plots the "echo" deeper than it should

Mirror Image


Mirror Image

· The biggest hazard with mirroring is that it produces the appearance that something exists where it does not · "Real" echoes are stronger, and may overwhelm the mirror effect · A real object may also make the interface less reflective so you only get a partial mirror


"Lobe" artifacts

· Side lobes and grating lobes are sound waves that are created peripheral to the main beam · Are lower energy, so generally are lost to attenuation · Provide distracting echoes when misinterpreted as being from the main beam


Slice thickness

· Remember that the image is a two dimension representation of a volume of tissue, not a true two dimensional plane · Results in false echoes, especially in anechoic structures

Slice thickness

· Thick slices result in an inability to see objects substantially smaller than beam thickness · This is also referred to as elevational resolution · Generally not controllable by the user · Generally the poorest resolution in the system


TGC banding

· Occurs when the user has not appropriately adjusted the TGC curve


· Occurs only in Doppler imaging · Happens for the same reason that wagon wheels appear to turn the wrong way in old movies · The computer assumes a baseline and a range (determined by PRF) · If the velocity exceeds the range, it will be depicted as going the other direction



· Can control aliasing by increasing PRF or changing baseline · It's important to correct in pulse Doppler so accurate measurements can be made · It's important to correct in color Doppler because it is easily mistaken for turbulence


· It IS possible to have the PRF set too high · A certain percentage of signal right around the baseline is discarded · When you are looking at high-flow structures, this throws out "garbage" movement like that associated with respiration or the pulsation of vessel walls


Angle correction

· Remember the Doppler equation?

­ Frequency shift =2f(/c)cos

· The computer does not know the incident angle unless you tell it · It will assume an angle of 0° in its calculations · We very, very seldom interrogate vessels at this angle


· Angle correction does NOT need to be performed if you are going to plug the velocity measurement into an equation where the units drop out

­ e.g. Resistive index (RI) = (S-D) / S

­ S=peak systolic velocity ­ D=lowest diastolic velocity




30 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate