Artifacts in Ultrasound Imaging - SD

Artifacts in Ultrasound Imaging

Artifacts are everywhere in ultrasound imaging
and are produced in three basic ways.
First, the machine must make simple assumptions in order
to create the image.
Second, the acoustic properties
of the tissues create artifacts within the image itself.
And finally, the physical characteristics
of the sound beam contribute to some of the artifacts
that you will see every day in ultrasound imaging.

We are going to take these three causes,
the simple assumptions, the interaction of sound and tissue,
and the sound beam,
and look at the artifacts
that are caused within each category.
First, we'll look at some
of the simple assumptions the machine must make in order
to create an image.

Simple Assumptions the Machine Makes

There are fundamentally six
such assumptions.
The first is that sound travels only in straight lines.
The second is that echoes arise only within the main beam.
Third, that the echo amplitude is determined
by the object alone.
Fourth, that the depth of the interface can be calculated
from the time that it takes for an echo to return.
That is a uniform velocity in soft tissue can be assumed.
Fifth, that soft tissue motion can be discriminated from
blood flow on the basis
of the color doppler signal in the gray scale.
And finally, that all registered frequency shifts
represent motion.

These are the artifacts that are produced from these
simplistic assumptions.
The first is reverberation, short path,
reverberation mirror image, multi-path artifact, side lobe,
focal zone banding flash,
and some newcomers to the artifact world to twinkle
and quiver.

Reverberation Artifact

Reverberation is created when sound generated at the
transducer face, hits a reflector
below the surface of the body,
and the sound is reflected back to the transducer face
and makes multiple round trips.
Each round trip is then transcribed
as a line within the image.
As a result, multiple straight lines can be seen in the near
field or spurious echos can alter the echo texture
of the image tissues.

Here is an example, a reverberation artifact.
In the gallbladder you can see three straight lines within
the cyst of the gallbladder, which is produced
by multiple round trips
From the transducer face to a sub reflector.
The absence or presence
of reverberation can be used diagnostically.
On the left you'll see an abnormal fetal skull.
On the right there is the normal fetal skull.
Notice that on the right, the normal fetal skull
creates reverberation artifact, which obscures the superior
or non-dependent portion of the brain.
No anatomic detail is seen on the left.
The anatomic detail is seen extremely well.
This suggests that the skull on the left
is under mineralized, as can be seen
with osteogenesis imperfecta,
hypophosphatasia and other conditions.
In fact, the child on the left did have
osteogenesis imperfecta.

Short Path Reverberation Artifact

A similar artifact is something called short
path reverberation.
This goes by other names
and has been referred to as comet tail
or ring down artifact.
Fundamentally, what happens is
that sound produced at the transducer face enters
and is trapped within some sort of reverber chamber
or some sort of enclosed space.
As a result, the sound bounces
around within this enclosed space
and sends back closely spaced echoes to the transducer, each
of which is written as a line of sound within the image.
One of the most common causes
of this short path reverberation artifact is air.
Presumably sound is trapped either within the interstices,
between the microbubbles
or within the microbubbles themselves.
And as a result, The ring down artifact
that you see on the image on the right is produced.
In this case, the bring down artifact is seen
behind gas within the duodenum
and is shown in the image as a kind of shaft
of high amplitude signal.

This artifact two can be used diagnostically
when a similar sort
of comet tail artifact is seen within the gallbladder.
And this is the gallbladder on this image here.
And here is the short path reverberation between
behind echogenic foci that indicates
that you have cholesterol crystals within the gallbladder.
Likewise, short path reverberation seen be
behind echogenic reflectors in the thyroid.
Ma uh, thyroid nodule indicate that
what you are looking at is a colloid cyst.
And finally, you can see this short path reverberation
illustrated here
behind calcifications within the testicle indicating
testicular myotis.

Mirror Image Artifact

A third artifact produced by the simplistic assumptions
by the machine is mere image artifact.
Mere image artifacts are produced when sound
follows A course to a secondary reflector.
It balances off, hits a, an additional reflector
and retraces its path back to the transducer.
The machine will assume the sound travels in straight
lines as a result.
The second, the secondary reflector.
In this case, this white circle
is projected in a straight line here at a spurious depth.

The classic example of this is the human genoma in the liver
that sits just below the diaphragm,
the diaphragm being the very echogenic, uh, reflection
and creating an epigenic reflection.
And as a result, the hemangioma is written
in a straight line above the diaphragm in the lung.
As I say, the machine logic assumes
that all sound travels in a straight line
and the time equals distance.
Therefore, the secondary reflector is written posterior
to the primary reflector.
In this case, you have a sound coming down,
hitting a primary reflector, then this genic mass
and then making a return trip.
And so you have it written here above the primary reflector
or the mirror plane, if you will.

Miriam, major facts are seen in many contexts
is seen here in this, uh, testicular examination
with a large hydros eal.
Here's the testicle right here.
And you can see the mi image artifact has created a spurious
appearance of a secondary test is below the level
of the scrotal skin here.
Secondly, laryngeal rings can be projected, uh,
at spurious depth below the air column
as seen on the image on the right Mi image.

Artifacts are not limited to the gray scale,
but can be seen in color doppler imaging as well.
In this case, we have images
of the carotid bifurcation here.
There's a brightly echogenic layer
of calcium here on the posterior wall of the
external carotid artery.
As a result, sound comes down,
hits this bright reflection, bounces up,
hits a moving blood cell and makes a return path.
And so this area of doppler signal is written
to a spurious depth here.
Notice how the color saturation
and hue closely resemble that of the reflected artery.
Here is the area of higher uh, velocity,
and that's reflected down in this phantom
or mirror image artifact.

Multi-Path Artifact

Artery multipath artifact
is quite similar to mere image artifact,
but the concept here is
that sound generated at the transducer face hits a primary
reflector and then multiple secondary reflectors on its
return to the transducer face.
As a result, you'll get multiple scattering events
as the sound echo returns to the transducer face.
This kind of artifact is responsible for the clean
and dirty shadowing that we see
Behind gas or behind calcium.
Say in the gallbladder.

Here's an example of a gallbladder with a large stone.
Notice that the surface here is curved and
because it is a stone is rough.
As a result, the re the sound beam
that comes down hits the primary re reflector, which is the,
uh, leading edge of the stone.
And the reflected beam is diffused,
it becomes weak and incoherent.
As a result, very few secondary scattering events occur,
and little of the of the reflected echo returns
to the transducer face.
As a result, the shadow behind the gallstone is clean.

Compare that to this case.
Here we have an epigenic reflector,
but behind it there is acoustic noise,
the so-called dirty shadowing.
In this case, what we have is air within the gallbladder in
the case of emphysematous cholecystitis.
And as a result, this ver the reflected signal is very
strong and coherent
and is involved with multiple secondary scattering events
on its way back to the transducer face.
Consequently, all those returning echoes are written into
the image as acoustic noise.

Side Lobe Artifact

Side lobe artifacts are produced
because the main beam is not the only beam producing
echoes within an image.
There are several smaller lower intensity beams
that flank the main beam.
These are sometimes called side lobes or grating lobes.
So this would be the main beam
and these are the lower intensity side
lobes or grating lobes.
These do not cause much problems in routine imaging except
where there are cystic structures.
These koic structures can
show evidence of the side lobe artifact if the side,
if the low intensity side lobes encounter a brightly
echogenic reflector.

The classic example of this would be the gallbladder.
The gallbladder which is koic
and filled with bile, Sits right next
to the duodenum, which is often has air in it
and causes strong reflection.
As a result, the si
of lobes encountering the air in the duodenum, uh,
will produce a strong enough reflection
that can be seen within the a koic background
of the gallbladder.
The side lobes hit the duodenal, uh, gas, it returns here,
but all those returning echoes are going
to be written into the territory of the main beam.
As a result, they're gonna be written over here.
The machine can assume only a single coherent beam
and cannot account for all the reflections that are caused
by these side lobes.
As a result, low level echoes from adjacent reflectors can
be placed within organs and cysts.

Here is an example of a side lobe artifact within the
gallbladder that looks like sludge.
This is caused by the highly echogenic
gas within the duodenum as demonstrated here.
There are some clues to this.
First, the artifact tends to have a con cave
or meniscal like appearance on the, on the gray scale.
And secondly, a bright reflector is seen in direct proximity
to the side lobe.
You can, you know, this is an artifact
because it is inconstant, it changes
from a different perspective.
So if you change the patient position
or you change the tra transducer position,
the side lobe will go away.

Focal Zone Banding Artifact

Focal zone banding,
the ultrasound machine must assume that the echo
or signal amplitude is a consequence
of the tissue itself and not other things.
In fact, the amplitude
of the returning signal can be influenced
by the coherence of the beam.
The beam is most coherent at the focal zone and
consequently the uh,
returning signal will have a higher amplitude
from the portions of the tissue that are intonated
by the at the focal zones.
As a result, you will get echogenic bands
produced at the focal zone.
Here's the transducer, here's the focal zone
where the beam is most coherent
and you will have brighter echoes from this area.
This is often seen in small parts imaging.
In this case we have a linear transducer intonating.
The testis, the focal zone is right here
and you can see a brighter band of echogenicity here
relative to the near field and the far field.
And that lighter band is, uh,
at the focal zone.
This is usually not a problem
unless the bright band isolates a segment of the testis
like this here in the near field, which can look
somewhat like a tumor.
So you can get a hypo coic area within the testes, which is
artifact and not
real.

Flash Artifact

Flash artifact is a result of the fact
that all motion will be color enc coded,
not just blood flow.
If a patient is breathing, you will get a flood
of color from the moving tissues.
Now the soft tissue color can be suppressed
where the gray scale echoes are detected,
but written only in hypo coic areas.
This is part of an imaging algorithm that can be selected,
is part of the machine presets.
The result is that you can get
Apparent flow within hypo coic areas
where no flow in fact exists
because color will be written preferentially only
to the darker areas.

Here is an example where this artifact led
to a misdiagnosis.
This was a child with cystic fibrosis
and advanced lung disease and advanced liver disease.
As a consequence, the patient was breathing rapidly
during the entire examination imaging,
the epigastrium reveals multiple hypo coic structures, some
of which had the appearance of tubular structures
color was put on, and you could see that the hypoechoic
or the IC areas filled with color suggesting
that flow was within these tubular structures.
And further suggesting that what we are seeing was
epigastric es from the advanced liver disease
and portal hypertension.
In fact, when the patient had a ct,
what we saw was the multicystic variety
of change within the pancreas,
sometimes seen in cystic fibrosis,
and that these were not vessels at all.
Rather, the flash effect
produced an apparent color signal within the hypo coic
structures, merely
because the color was suppressed over the gray scale,
but allowed to be written in koic areas.

There are clues to this.
First, if you put spectral doppler in the area of the color,
a color aberration,
you will not get a coherent spectral doppler signal.
It will not look like an artery or a vein.
Secondly, this is often seen when patients are unable
to suspend respiration.
Third color sensitivity is usually relatively high and a
and a color preset is such that color is written only
to the darker portions of the gray scale.
And finally, you will get unusual color patterns, patterns
that don't really make any sense single color
with an entire region or color bleeding.
You will not see, uh, blood flow that appears
to come in one side of the image and out the other.

Twinkle Artifact

A similar artifact is the twinkle or quiver artifact.
This artifact has been more recently described as a band
of color behind a highly reflective interface.
As demonstrated here, this color is generated
behind the interface of a renal stone.
The explanation for this artifact, uh,
lies in the recognition
that the ultrasound machine has a digital clock within it
that is responsible for the generation of the pulse
and the timing of the echo.
There is always some amount
of variability in this clock at some level, so you can get
what is called phase
or clock jitter,
which is essentially just noise at the level of the
silicon electronics of the machine.
These tiny time errors produce slight errors in phase
and frequency shift and produce doppler noise.
The noise spectrum is magnified
when you have a bright brightly reflective interface
and a roughened surface
because there is increased discrepancy
between the path lengths of the transmitted
and received signal.

Here is the example of what I was just describing.
The path length air is interpreted as face ship shift,
which is the same motion.
So if you have a roughened surface such that sound counters,
uh, a higher peak
or a valley, that
that you will get a certain dis definable phase error,
which will be seen as motion by the machine.
By the machine.
This noise then exceeds the limits of the wall filter
behind these highly reflective interfaces.
And so you'll have a situation like this.
This is a kidney with an epigenic stone.
If you turn on the color, you'll get this bright band
of color produced by the, uh, phase jitter
or the twinkle artifact.
And if you put a spectral doppler tracing in there,
you'll see that this is high amplitude noise
and not a coherent doppler vascular signal.

This can be used diagnostically.
Sometimes small stones are difficult
to identify either in the kidney, in the ureter
or in the gallbladder,
but the presence of twinkle artifact can
clearly identify their location.
Here's an example of a large stone
seen within the gallbladder,
and you can see the twinkle artifact, uh,
behind the echogenic leading inter leading
interface of the gallstone.
This is of course not true flow,
but represents the twinkle
artifact produced at that interface.

Quiver Artifact

A similar type artifact is what we have called here
in our institution, quiver artifact.
Quiver artifact is similar to twinkle in that you get this,
uh, artifactual doppler
or power doppler signal at the location
of brightly epigenic interfaces.
In this case, however, the artifact is not caused by the
acoustic noise of the machine itself,
but rather from micro movements, uh, produced
by acoustic radiation force.
We've seen this in presence of stents.
For instance, this is a stent within a patient
who had an occluded tips stent
and there was no flow within the stent.
But at the stent wall we get this powered doppler signal.
Similarly, we can get this signal within a stent
buried within a phantom.
When you do an experimental study on this, you can do,
you can demonstrate displacement from frame to frame,
and here is the power doppler signal.
You can see the movement here frame by frame
of the stent within this phantom.

So the responsible agent
for this quiver artifact is acoustic radiation force.
Remember, the sound waves are pressure waves
and a force is imparted
by the pressure gradient of the sound pulse.
The momentum of the pulse then is transferred
to the medium causing small tissue displacement.
Little micro movements
and the images created from these displacement estimates
is demonstrated
as doppler signal along brightly echogenic interfaces.

Artifacts from Interaction of Sound and Tissue

A second group of artifacts are produced by the interaction
of sound and tissue.
These are the artifacts that I'll discuss.
Speed propagation artifact, posterior shadowing,
posterior enhancement refraction, phase cancellation
attenuation, perivascular color artifact,
and a artifact known as an isotropy.

Speed Propagation Artifact

Speed propagation artifact is produced
because the machine assumes a uniform velocity
of sound in soft tissue, specifically that
of 1,540 meters per second.
This estimate is good for all major, uh, soft tissues
except for fat.
You can see that the difference between this
soft tissue average of 1540 and the speed of sound
and fat is much different than the other soft tissues.
As a result, sound that travels through a fatty mass
or a collection of fat will be slowed in its passage through
that mass and it's also slowed in its return passage through
that fatty mass from a sub adjacent reflector.
So the time of echo return is equated with depth sound
like arising from the transducer is slowed by this area,
this fatty mass here, which sits above the kidneys,
it'll be slowed in its passage to the diaphragm
and it'll be slowed on its return.
As a result, the diaphragmatic echo appears to be disrupted
and you'll see a step off right here.
The diaphragm, of course, is intact,
but what we're seeing is a speed propagation artifact which
makes that appear to be discontinuous.
In that case, that fatty mass was a myelo lipoma
of the adrenal gland.

Posterior Enhancement Artifact

Posterior enhancement sound is attenuated
to variable degrees by different tissues.
Fluid, fluid attenuates sound less than soft tissue.
As a result, the more energetic beam
produces higher amplitude echoes behind a cyst,
so tissue intonated
by a more energetic beam will appear more
energetic, more echogenic.

Here is an example. Sound produced at the transducer
interface is attenuated at a regular rate until we get here
to the cyst where it's attenuated to a lesser degree
than sound in the soft tissue just adjacent.
As a result, you have higher amplitude echoes produced
behind the cyst, the so-called posterior enhancement, uh,
this sort of flame, if you will, of echogenic tissue
of agen genicity behind the cyst.
This, of course, is used diagnostically all the time
to identify cysts.

Note, however, in this case, on the right
of an ultrasound image of an ovary, you get no acoustic
enhancement or posterior enhance
behind this apparent cyst of the ovary.
The reason for that is that that is not a cyst
but a fibroma of the ovary.
And as a result, the absence
of posterior enhancement a defies this not
as an simple ovarian cyst,
but rather as something quite different.
I will say as a caveat, as a word of warning,
that some very homogeneous soft tissues,
particularly lymph nodes,
can produce this posterior enhancement phenomenon.
So it is not unique only to fluid-filled structures.

Here's an example of a man with, uh,
a lymphomas mass here in the pelvis,
and you see this bright echogenicity behind the mass
suggesting that this is a cystic mass,
but in fact, here's the CT demonstrating this
to be a collection of lymph nodes.

Refraction Artifact

Refraction artifact is produced when sound encounters an
interface between two media at an oblique angle.
There are two conditions that must be met
before you get refraction artifacts.
One, the two media must have different,
must conduct sound at different velocities.
And secondly, the encounter,
the sound beam must encounter the interface
between the two media at an oblique angle, so you get a kind
of bending of the sound beam here at that interface.
In fact, the sound beam tends to bend towards the media
of slower velocity as demonstrated here.

These are cis sound, uh, moves more slowly
relative to the sound here in this more solid, uh,
tissue in this phantom.
And you can see that the refracted sound shadows,
the refractive shadows tend to bend inward.
Compare that to this case
where the sound in the fetal skull tends
to bend the sound outward
and you get this sort of lateral displacement of echoes
and an apparent disjunction here in the fetal skull.

Anisotropy Artifact

The last of these artifacts produced by the interaction
of sound and tissue that I'll describe is called
an isotropy.
If you look at this word, you can see that it means an not
iso equal trope are turning.
So this is not equal.
The images are not similar from different perspectives.
Genicity changes with the angle of intonation
because the orientation of the tissue planes relative
to the sound beam axis changes.

Here is an example in the fetal head, I'm sorry,
the neonatal head.
You can see that we are imaging
through the anterior font nail
and getting the genicity here
in the white matter posterior to the ventricle.
This is because the white matter tracks are so oriented
that they are parallel
to the sound beam when inated in this region,
but perpendicular in this region.
As a result, there'll be greater reflectivity
from these perpendicular interfaces from these as compared
to these interfaces oriented parallel to the sound beam.

The way you can demonstrate that this is not a real finding,
that this is not true, echogenicity in that portion
of the fetal of the neonatal brain is
to change your perspective.
And here we have rotated the transducer
to the posterior il rescind this area.
Now the the orientation
of these white mi fibers is in parallel to the sound beam
and the increased echogenicity seen in the periventricular
white matter has disappeared.

And isotropic is also seen in other areas
where there's a regular repeating pattern of interfaces such
as in the pylori of the, uh,
of the infant with pori stenosis.
Here you can see the echogenic area reflect, uh,
indicating the mucosa,
and here is the hypertrophied muscle in, uh, pric stenosis.
If the sound beam comes down
and encounters the muscle, uh, perpendicular to the
axis of the sound beam, you will get greater reflectivity.
Then when it encounters the muscle, uh, fibers
that are parallel to the sound beam, as a result,
the echogenicity here in this portion of the pylorus
will be greater than that here on the sides.

Summary and Quiz

So, so artifacts are produced by a number
of different causes.
It's produced by the simplistic assumptions the machine must
make in order to create an image.
It is produced by the interaction of sound and tissue
and is produced by the sound beam itself.
Now we'll have a summary and quiz.

Take a look at this image on the left
and see if you can tell what has caused this artifact.
Well, this is sometimes called the renal
duplication artifact.
This is the, this is the kidney
sound coming down can be bent by the interface
of the fat of the peri hepatic or per splenic fat.
And as a result it's refracted or bent down,
but it will be written in a straight line.
So you have what appears to be two poles,
two upper poles of this kidney.
So this is a refraction artifact.

Take a look at the image on the right,
what artifact is illustrated.
Here again,
you see an echogenic mass sitting above the kidney
Being genic suggests that it might be composed of fat,
and this is confirmed by this disjunction
or apparent disjunction of the diaphragmatic echo here.
This is a consequence of a speed propagation artifact.
Sound originating at the transducer is slowed
through this fatty mass both coming and going,
and as a result, the round trip of sound back
to the transducer will be slower than it will be in the
portions of that travel through the liver only.
As a result, you will get a displacement of the
diaphragmatic echoes to a spurious
or a, um, uh, erroneous depth.

What artifact is being illustrated here on the left?
Well, you, your imaging the epigastrium of an infant.
You can see that the area of the pylori is thickened
and then there's this genicity here
as the muscle bundles are oriented perpendicular
to the sound beam and a decreased echogenicity
as the muscle bundles are oriented in
parallel to the sound beam.
This, of course, is an isotropy.

Here's another artifact illustrated.
This is an image of the thyroid gland.
This is the thyroid here,
and in the thyroid is a nodule within the nodule,
our echogenic foci with little tails of echogenic
genicity behind them.
This is the so-called come tail artifact
or short path reverberation sound comes down
and is trapped in these reverberation chambers
and comes back and the returning echoes are
therefore at very short intervals,
very closely spaced intervals,
and you get this little tail behind them.
This indicates that what you,
what you are looking at is a colloid cyst within
the thyroid gland.

Look at this artifact.
This is a scrotal examination, large hydros eal testis here,
and you can see behind the scrotal skin there's another
testes in more hydros eal.
This is because you have a mirror image.
Artifact sound is being reflected along this
mirror plane, which in this case is the scrotal skin
and creating the appearance of two testicles
in the same image plane.

Look at this artifact here.
In this case, we are looking at the gallbladder.
This is the gallbladder here,
and within the gallbladder are these very echogenic foci.
Behind the echogenic foci is a lot of signal
or dirty shadowing.
Such dirty shadowing is produced
because the sound reflected from the gas.
In this case, gas is involved
with multiple secondary scattering events
as it returns to the transducer,
and all those returning echoes are gonna be written into the
territory behind the reflective gas.
This dirty shadowing is a consequence of what is known
as multi-path artifact.

Look at the case on the left, you can see
a kidney here with color.
You can see many vessels coming into the kidney,
but then there's this bright band of color right here.
That band of color is produced behind the interface
of a renal stone, and this artifact has been referred to
as twinkle artifact.
Remember that this is noise produced at the level of
the, uh, machine is sometimes referred to as phase
or clock jitter,
and it is seen only with color doppler signal
because that phase shifted, that phase shift,
that erroneous phase shift produced at the rupin surface
of a brightly reflective interface is gonna be interpreted
as doppler shift and written as such.

The last review is this, this is an image of a testicle.
Here's the testicle here,
and you can see in the middle
of the testicle is this brightly reflective band right here,
and then there's a band of lower genicity,
both in the near field and in the far field.
That brighter echogenic band is produced at the focal zone,
and this artifact is known as focal zone banding.

Well, this is a collection of artifacts that you will see
routinely in in ultrasound imaging.
As you become more familiar with them.
You can be these, you'll recognize
that these artifacts can be used diagnostically
and also their misleading effects
can be seen through as well.
Thank you for your attention.