Contrast Enhanced Ultrasound in the Abdomen - SD

Introduction

Hello, I'm Bill Lees.

I'm a professor of radiology at University College London,
and I work at University College Hospital in London.

I've been doing ultrasound since 1974,
and I'm going to talk to you about
contrast ultrasound in the abdomen.

Ultrasound contrast is licensed in the UK
and I'm going to show you how we use it in our
regular clinical practice.

Contrast Enhanced Ultrasound in the Abdomen

One of the things we've discovered in using contrast over
the last 15 years is
that ultrasound is inherently non-linear.

Many of you'll be familiar with tissue harmonic imaging
where the property of ultrasound that
various frequencies do not travel at the same speed
through tissues, leads to the generation
of harmonics within the tissues,
which are radiated back from the tissues to the transducer,
giving us different frequencies to the ones
that were actually input into the tissues to start
with, it's a bit like the same phenomenon of a wave
as it approaches the shore, that wave gets steeper
and steeper, and eventually the wave breaks
and creates noise.

Those are the harmonics
that we're imaging in tissue harmonic imaging.

It's an inherently non-linear system,
and that's important to remember.

We use this for contrast imaging today using the technique
of pulse inversion, where in a perfect system,
if we take an acoustic pulse here, echo number one,
and then input a second pulse,
which is exactly 180 degrees out of phase from the first one,
if we add those together,
then we get absolutely no signal returned.

They cancel each other out perfectly.

And we use this with contrast imaging
to suppress the background echoes from normal tissues.

In the case where we have
harmonics generated by the bubbles
or within the tissues, this summation process
does not work perfectly,
and we are left having canceled out the tissue signal
with harmonic signal enhancement, which is coming purely
from the contrast agent itself from the bubble.

And it's exploiting the non-linear property of the bubble
that it does not reflect the sound that was input,
but the harmonics and other signals generated
by vibration within the bubble itself.

This has been extended into techniques varying from
manufacture to manufacture.

One is called power pulse inversion, where instead
of just using two pulses to cancel out the signal,
multiple pulses are put in,
which can also correct for tissue motion.

And these are displayed as a color overlay,
with the fundamental tissue image behind.

Here's an example. You can see
multiple signals put in all at slightly different phases,
when they're summed from the tissues even moving
tissues, the background signal is suppressed so
that all we're left with is signal from
the contrast agent itself.

So here's an example where we're superimposing the contrast
signal over the B mode image
and showing with very, very high contrast.

A small lesion here deep within the liver.

Ultrasound Contrast Agents

Ultrasound contrast agents are all very similar.

They're stabilized microbubbles.

The optimal microbubble size is one
to three microns in size.

If you think of a red blood cell is seven microns in size.

So these are significantly smaller than a red blood cell.

They can pass through the pulmonary circulation.

So they can be given by intravenous injection rather than
intra arterial injection,
and then supply the whole of the systemic circulation.

But being of the same order of size as a red blood cell,
they don't escape through the vascular endothelium
into the extracellular fluid compartment.

They can find purely within the vascular space.

Blood pool agents, the microbubble structure usually consists
of a phospholipid shell, initially containing
air in the first generation of contrast agents.

But now they can contain fluorocarbons
and similar compounds, which are more stable gases,
which don't dissolve readily in the serum
and persist within the circulation for a number
of minutes up to 10 or 15 minutes.

With some of the later agents.

There are a number of different agents available
in the market.

You can see here, and again, an example of red blood cells,
with the microbubble contrast agents
scattered throughout them.

Typically there'd be a ratio of about one
contrast bubble to a hundred red blood cells,
and this is enough to get a signal from the blood pool
that we can image a single bubble under optimal conditions.

There are a number of different agents,
which have different compositions.

They've been the first generation of contrast agents
were operating at a high mechanical index,
often with bursting of the bubbles
to produce the harmonic signals.

The current agents are sufficiently stable
that we can image them at very low power,
which is low mechanical index,
and we can scan them for long periods
of time without destroying the bubbles.

And we can visualize them in different phases,
just like in CT or MR contrast studies with an arterial
and portal venous phase,
and even a late parenchymal phase
where we're looking at bubbles
that are trapped within the micro vasculature of the tissue,
which is a unique property of ultrasound contrast.

Something that we can't do with CT or MR
contrast agents currently.

So if we look at the microbubble, when it's intonated
with an acoustic wave with a positive pressure
and a negative pressure phase, the typical wave form of
longitudinal acoustic wave in the high pressure phase,
the bubble is compressed to a very small size.

And in the low pressure phase,
it will expand quite dramatically to a much larger size.

And these bubbles will vibrate during this process.

And it's this vibration which generates the signal harmonics
that we're imaging to obtain the characteristic signal
that comes purely from the microbubble itself.

So here we see the nice clean input signal.

This is the appearance of the signal plus the harmonics.

When this is fur transformed, we get the various
harmonic signals generated.

Here's the fundamental frequency.

This is the sub harmonic at half the frequency,
twice the frequency three times, and so on.

And we can separate out these signals using filtering
techniques really very effectively today.

Low Mechanical Index vs High Mechanical Index

So a low mechanical index,
this can be anything from 0.01 to 0.2 compared
with a normal mechanical index of an ultrasound scan,
which is the order of one to 1.5.

The bubbles are preserved, so we can use real time scanning.

We do need a sophisticated ultrasound machine to do this.

But these are plentiful in this current day
and age at high mechanical index, for example, 1.2.

So the normal standard ultrasound imaging power,
the bubbles will burst.

These produce a very bright flash of an echo
as each individual bursts, which is very easy to see
even on quite basic equipment.

But we need to use special intermittent scanning techniques,
having burst the bubbles as we wait
for the bubbles to replenish.

This in itself can be a useful property if we purge
the image field from bubbles
as the bubbles come back into the field from
outside and replenish.

We're actually visualizing perfusion at the tissue level,
which is again, another unique property
of ultrasound contrast agents.

High mechanical index gives us any limited
views per injection.

Obviously the technique of scanning is quite difficult,
and currently the first generation agent Levovist,
which is still licensed in Europe and
most of the world, is the optimal agent
for this bubble bursting technique.

But it's been superseded by the second
and third generation agents, such as
SonoVue, Optison and so on,
and many that are licensed in some countries,
many not licensed in the continental United States.

Clinical Use in the Abdomen

So how do I use ultrasound contrast within the abdomen in my
day-to-day practice, where
I'm using fully licensed contrast agents in the liver,
the main use is in focal liver lesions,
obviously detection.

We can use contrast agents just as we do in CT
or MRI to improve a detection of metastasis
and primary liver lesions.

If we pick up liver lesions on a conventional ultrasound,
either in a screening context
or as incidental findings, we would like to be able
to characterize those lesions without having to go on
to do CT or MR scans.

So, if we can do as well with ultrasound contrast
as we can with these other techniques,
then we can save significant sums of money.

One of the important things in planning surgery
or radiofrequency ablation is to determine the exact number,
position and extent of the liver tumors
so we can plan our therapeutic procedures.

And also we can use contrast agents to evaluate
before we do ablation procedures
to monitor the ablation procedure itself
and to follow up the result of ablation.

And this is quite routine in my practice in my hospital
where we're doing seven or eight radiofrequency
ablations each week.

And there may well be a role in patients
with diffuse liver disease, such as cirrhosis
or HCV and HBV infection, where we can screen
the liver more effectively for early detection
of primary hepatocellular carcinoma.

Focal Liver Lesions: Hepatocellular Carcinoma

Now, if we're looking for lesions such
as hepatocellular carcinoma, as you will know from CT
and MR Imaging, we are looking very much
for the vascular characteristics
of these particular lesions.

As you know, a hepatocellular carcinoma will tend
to show a strong enhancement in the arterial phase
with early washout in the portal venous phase,
and this is due to a fundamental property of the liver
with its dual arterial and portal venous blood supply.

And the transformation of normal liver tissue
to tumor tissue involves the progressive switching
of blood supply into the art pure arterial phase
and away from portal venous supply.

And this is progressive as we go through from normal liver
to a regenerative nodule.

We have the normal balance between arterial
and portal venous supply with perhaps 70%
of the normal liver tissue being supplied from
the portal venous circuit.

We go through dysplastic nodules.

We gradually introduce an abnormal arterial supply,
and as we go through progressively
more malignant lesions from well differentiated
to poorly differentiated HCC, in the end,
this transformation is complete
and the poorly differentiated HCC will derive its entire
blood supply from an abnormal arterial circuit
and almost none from the portal circuit.

This property was used extensively in the early days of CT
where we were doing CT angiography,
using direct hepatic artery injection of contrast,
and then portal venous imaging using injection
of the superior mesenteric artery to fill
the portal circuit after circulation through
the intestines.

And it was shown very elegantly back then that
these lesions, even quite small lesions, if they were
truly malignant derived almost their entire blood supply
from the hepatic arterial circuit with minimal supply
from the portal venous circuit.

This involved arterial cannulation when transfer
to the CT scanner and rather fell out of favor.

And we can get almost as much information today
with modern CT scanners from a rapid intravenous bolus,
although there is some blurring between the two,
the arterial and portal venous phases.

But with practice diagnostically,
we can do almost as well.

Interestingly, the CT arterial portography is coming back
into fashion, where particularly interventional radiologists in the far
East are working with hybrid systems, combined angiography
and CT systems, particularly where they're performing
super selective embolization and chemo embolization.

So this technique hasn't gone entirely out of fashion.

It is coming back quite well.

Now that property can be visualized using ultrasound
contrast agents almost as well.

Here's an example of a large four centimeter
HCC within the right lobe of the liver visualized
by conventional B mode imaging.

Here is the arterial phase of a
ultrasound contrast study given by intravenous injection.

We can see the feeding vessel here into the tumor
with large vessels in the center of the tumor,
and a very strong arterial phase enhancement
with minimal enhancement of the lung parenchyma around
fairly characteristic features of an HCC.

And here we can see the same sequence again,
with the conventional B mode image
before contrast injection strong arterial phase enhancement,
here with some central necrosis.

And here you can see this what is called early washout
wherein the portal venous phase at about 60 seconds
after the intravenous injection.

We have relatively poor enhancement of the tumor
with strong enhancement
of the surrounding normal liver parenchyma.

According to the criteria
of the European Association for Study of the Liver,
visualization of this classic pattern is sufficient
to diagnose a hepatocellular carcinoma
without the need for biopsy.

Metastases and Other Liver Lesions

We can also use the same techniques for diagnosis of
very small metastasis.

This is a colorectal liver metastasis.

It's about a centimeter in diameter here in the portal
venous phase, and we can see good enhancement
of the normal liver parenchyma around,
with very high contrast ratios between the tumor
and the surrounding normal liver.

This is the corresponding CT scan
where we can see this small tumor,
which has an enhancing halo.

We're more in the arterial phase here than in the portal
venous phase, but you see exactly the same
temporal characteristics with contrast ultrasound
as you do with CT.

We can also have some interesting observations in chronic liver disease.

Here is a patient with cirrhosis, with nodule formation
that on the basic B mode image, we can see a rather
diffuse nodular pattern.

After administration of contrast
and viewing in the delayed phase, we can see
that this nodular pattern has been enhanced,
and we can see much more clearly
the individual nodules their size
and size variation that go up to make the
standard cirrhotic pattern.

So we can use this in diffuse liver disease, not only for
detecting malignant focal lesions,
but also for enhancing the pattern of nodule formation.

Some people have used contrast
to look at physiological parameters,
and one of these is the mean hepatic transit time
where one is looking at the difference in time
between the arrival of contrast in the hepatic artery
and portal vein to the arrival
of contrast in the hepatic veins.

And it's known, been known for many years
that this is shortened in cirrhosis
where patients have a hyperdynamic circulation.

This actually compares two different contrast agents,
but it does show that it is possible
to obtain physiological information if
that's what you're interested in.

By using contrast agents.

We have used HI MI imaging using the first generation contrast agent
Levovist for studying other types
of tumor within the liver, particularly Klatskin tumors
or hilar cholangiocarcinomas,
which can be extremely difficult to see
by conventional ultrasound by CT or MRI scanning.

This is using the delayed high power
high mechanical index technique.

And you can see here an abnormal texture
in the central portion of the liver
on the pre contrast scan.

But here, looking in the delayed phase
after contrast, we can very clearly demonstrate the full
extent of this malignant process
within the central part of the liver.

Here's another example showing very clear delineation
not any of the full extent of the tumor at the hilum
of the liver, but looking at the very sharp boundary
that this tumor presents, we can see whether
or not there is direct invasion
into the surrounding liver parenchyma
or merely a simple mass effect.

You can get the same information occasionally
from CT scanning.

Here's an example of a nice 3D study,
where we have particularly clear visualization
of the tumor mass at the hilum of the liver.

But this is unusual for CT
and what we found in our studies of this process,
that the ultrasound contrast technique is as effective
as the best technique we otherwise have, which is MRI
and significantly better than CT scanning at demonstrating
the presence of a mass that demonstrating the size of a mass
and demonstrating the extent
of invasion into the surrounding liver parenchyma.

Very often the clearest images are obtained
by ultrasound contrast rather than by CT
or MRI though each have their own particular advantages.

Now, there've been a number of studies done in Europe
and in Asia using ultrasound contrast looking
for liver metastases.

This is a study in which I took part,
looking at the first generation agent
and comparing with spiral CT.

And what this study shows is that
the pulse inversion contrast imaging at ultrasound
is significantly better than conventional ultrasound showing many, many more lesions
and is also capable of performing almost identically
to what was then state-of-the-art CT imaging.

Obviously, CT imaging has improved since then,
but so has the ultrasound contrast imaging.

So as is typical in radiology, all are imaging techniques
improved in concert and remain
equally competitive at all times.

And there are many studies
that have confirmed this contrast.

Ultrasound is at least as good as CT scanning
and is probably nearly as good as MR scanning in the liver.

Now, one of the important areas in the liver that we need
to establish, particularly in patients with
colorectal cancer liver metastases, is the presence
of very small lesions or satellite lesions.

And this is what I mean by a satellite lesion.

Here is a large colorectal cancer,
which has been surgically excised,
and here you can see a sub centimeter lesion.

It is about one to two centimeters away from the main tumor
mass and colorectal cancer metastases spread
through the liver, largely by invasion
of the vascular spaces, and can grow long blood vessels
and appear as separate metastasis within this halo
of two or three centimeters around the main mass.

This has always been the reasoning
behind the surgical dictum that if you excise
a colorectal cancer metastasis,
you must take at least a centimeter of normal tissue
as a safety margin so
that you gather in these often microscopic satellite tumors.

And we know that CT
and MRI are not very good at picking up sub centimeter
lesions because the inherent contrast is
insufficiently great.

Here we can see in the same patient
a satellite metastasis gain well under a centimeter.

This is about five millimeters in diameter.

We can see again, it's about a centimeter
away from the primary tumor,
and we can see exactly
where this is in relation to the primary tumor.

So it's obviously important to remove
or ablate this lesion as well as the primary tumor mass.

Here's a second distant metastasis.

Again, you can see the very high contrast that we have in
contrast ultrasound between the poorly enhancing tumor
and the surrounding normal liver.

So we can be quite confident in diagnosis of lesions down
to about five millimeters in diameter.

There are various different modes
for displaying the ultrasound contrast result.

Here's another metastasis about two centimeters in diameter.

Again, note the very high contrast ratio between the tumor
and the surrounding enhancing normal liver.

This also applies
to looking at the periphery of individual tumors.

And one of the features we've noticed
with colorectal cancer metastases with CT is
that we often have arterial phase enhancement
on the periphery of the lesion rather than the center.

And we can see this pattern of these
separate tumor nodules growing out
or budding out from the main tumor mass very clearly
with contrast ultrasound.

And this type of pattern has been associated
with a poor prognosis.

If you look at the basic B mode image, those of you
who are experienced in ultrasound will recognize
that there's obviously an abnormality there,
but the clarity of definition of that mass
is dramatically better with the contrast ultrasound.

Here's another example of a paper published
in 2003 where we can see
contrast in ultrasound shows more metastases than CT.

And in patients who've had other
standard reference imaging techniques such as MRI,
intraoperative ultrasound
or examination of implanted livers, CT
does not perform as well as ultrasound.

And there are many papers to confirm this now.

So contrast ultrasound in detecting metastases is
as good as MRI.

It's better than CT scanning.

Another example, showing demonstration
of sub centimeter lesions in relation to
a large metastasis with an enhancing rim.

Very clear definition, very important for surgical planning.

So that's detection. It works.

Characterization of Liver Lesions

Characterization of liver lesions.

Well, the principles on which we work are very similar
to those of contrast studies at CT and MRI.

This is a study again from 2002,
using the high mechanical index technique,
which was all we had in those days, trying to compare the
properties of benign and malignant lesions.

And we can see that homogeneity on the late phase
is typically benign.

No late phase enhancement is typically
malignant and so on.

Here's some examples.

Again, an obvious metastasis very clearly seen.

Here's a tiny metastasis seen on the periphery of the liver,
which is less well-defined as CT equivalent.

Lesions such as focal nodular hyperplasia
do show characteristic features
and typically they will show strong lesion uptake
of contrast and late phase imaging,
which we don't see in the typical malignant lesions.

And in many cases, it's possible to see the scar,
the central scar of focal nodular hyperplasia,
which the pathologist tell us is always present,
but it may be no more than a millimeter or so in thickness.

So here's an example of focal nodular hyperplasia.

Not much to see on the conventional ultrasound image.

Here's the central scar
visualized in the arterial phase of this FNH.

If you see the central scar,
you can be very confident in your diagnosis.

The temporal enhancement pattern also lends
to significant confidence.

Being able to characterize hemangiomas at the time
of the original ultrasound scan is important
and can save considerable costs.

Typically we will see
a characteristic GLO enhancement on the periphery
in the arterial phase.

And this is particularly true when we're dealing
with larger lesions, which we are much more likely
to classify as a typical
from the morphological characteristics.

Small lesions we can usually characterize very well
just on their morphology.

It's the atypical lesions that are more difficult.

Here's an example. Here's a five centimeter diameter hemangioma.

It's characteristics are not typical morphologically.

It's not particularly genic
or well-defined hereafter administration of contrast.

We can see the absolutely classical
nodular peripheral enhancement that we're used
to seeing at CT.

We can say with great confidence that this is a hemangioma
and requires no further attention.

The low mechanical index technique,
it gives us a much greater opportunity to study blood flow
during the different phases.

In fact, we can scan continuously for several minutes through the volume of a tumor.

And hence, we're not looking at phases,
but a complete dynamic examination.

So, the other factor that's important with the low
MI technique, if you go back to one of my original slides,
the tissue harmonics that are generated by the passage
of an ultrasound pulse
through the tissue are reduced at low power.

They're really a function of high power imaging.

So, we still retain the bubble harmonics,
however, so we get a better signal
to noise ratio at low power than high power,
which is completely unlike CT or even MRI.

So at low MI imaging, the initial phases emulate dynamic CT.

We're looking at arterial portal venous imaging,
but the late phases,
the bubbles are retained at the microvasculature.

And this is closer to the kind
of reticular endothelial imaging we will do in the liver,
using specific ultrasound.

MRI contrast agents such as SPIO
and the agent that is licensed in Europe for this type
of imaging currently is the agent SonoVue,
which is the most widely used agent in
Europe at the present time.

And we use this almost exclusively in this low mechanical
index imaging mode.

Here's an example. In the early vascular phase,
we can see we are visualizing the individual
large vessels particularly well showing no enhancement
in this malignant tumor.

Here's another example of later phase imaging,
where we can see, again, the micro vasculature
of the tissues with very poor microvascular filling
in the tumor itself,

because the technique is very sensitive
to very small lesions.

Here's an example of a patient planned for surgery
for resection of metastasis,
and we can see in this CT scan
multiple enhancing lesions down
to a few millimeters in diameter.

Now, this is an exceptional CT scan.

We will only get this kind of image quality
possibly only for a few seconds during one part
of the arterial phase.

And in CT. Now we divide it up into early
and late arterial phases,
but we can never predict exactly which point in time
will give us the best image.

With the ultrasound contrast,
we can detect these lesions equally well,
but by continuous scanning,
we are less temporally rigid than we are with CT.

One of the factors that enables us to perform treatment so effectively in the liver is the use
of contrast here, for example, is an HCC immediately
after radiofrequency ablation, where we can see
a fairly poor definition of the tumor here on CT.

We can see preservation of a small area
of enhancement indicating residual tumor.

We can use this quite effectively
with ultrasound contrast to control the effectiveness
of our radiofrequency ablation
and pick up quite small nodular areas and determine whether
or not there is residual tumor which we need
to go back and to treat.

So in my practice now,
ultrasound contrast is fairly routine in liver imaging
wherever we're trying to detect
or characterize focal liver lesions,
and particularly for controlling biopsy
or radiofrequency ablation techniques,

Pancreas Applications

we can use it in the pancreas as well.

And here is an example of exquisite visualization
of a pancreatic tumor.

This is the corresponding MRI slice
where we don't see the tumor.
Anything like as effectively you can see on the MRI,
there is some thickening of the duodenal wall.

Here we see it on the contrast enhanced ultrasound.

This is the mucosa of the duodenum.

You can see the layered pattern,
which is enhanced normally,
and this is poorly enhancing tumor
invading the muscularis propria
of the duodenum in almost a circumferential pattern.

So we're seeing that really well both anteriorly
and posterior to the tumor.

But here's the key finding on this scan.

We can see a tiny tongue of tumor here growing out
towards the enhancing superior mesenteric vein.

And this ability to show the relationship of the tumor
to the portal vein is vital in planning the type
of surgery that we're going to perform for primary pancreatic tumors.

And in studies to date,
ultrasound contrast has performed at least as well as CT
and MRI in staging primary pancreatic carcinoma
and determining the vascular relations
and degree of vascular involvement.

Research Opportunities and Other Applications

Now, there's a large number
of different research opportunities with ultrasound contrast,
particularly when we're trying
to study animal models.

And one of the things that properties of ultrasound is
that it can be scaled up
or down very effectively in a way that is more difficult
to do with CT
or MRI In our small animal imaging laboratories,
we do have micro CT
and micro MR scanners for studying small animal models,
but these are extremely expensive, yet ultrasound can be
scaled down by increasing frequencies
and by building special transducers.

And here, for instance, you can see a Doppler study
of a two millimeter tumor in a mouse,
where we can enhance the Doppler study by contrast
to pick up what is really microvasculature within the tumor
and study the effects of things like
anti-angiogenesis agents
and see their immediate effect on tumor blood flow.

Here you can see a mouse heart being studied
with ultrasound, the mouse heart beats at between 300
and 600 beats per minute.

This is far too fast to be visualized by CT
or MRI, but ultrasound on this scale at least,
which is less than a centimeter, has the ability to keep up
with these incredibly high frame rates.

And ultrasound contrast, as you see here,
will not only demonstrate for us the chambers,
but here you see bubble bursting a high mechanical index pulse
to purge the myocardium of bubbles.

You can then look for reperfusion
and there you can see it occurring even on this millimeter scale.

And this is translated into myocardial perfusion studies in man,
where we can demonstrate the extent
of infarct really very effectively.

We can also image selectively the microvasculature.

If we're looking in the late phase after the initial arterial input
bubbles get trapped in the small capillaries
and micro vessels of a tumor.

And we'll very slowly pass through in any individual scan snapshot.

We may visualize a few of these small microbubbles,
but we don't get a picture
of the micro vasculature as a whole.

But if we have a very long persistence mode
and some images in sequence,
we can build up a gradual image of the full extent
of the micro vasculature.

Here you can see the tumor and necrotic center
and multiple feeding
and draining vessels in this small breast cancer.

And this technique of microvascular imaging,
or MVI, as some of the manufacturers call it,
is extremely powerful in imaging structures such
as the breast and the prostate.

It can be applied to larger lesions in the liver, again,
looking at late phase imaging,
looking at the smaller numbers of microbubbles
as they slowly traverse the lesion
so we can build up a pattern of both small
and large blood vessels within the tumor
and see exactly what is happening.

So now using these techniques, we're able to visualize
with contrast ultrasound blood vessels of the order
of 50 microns in size.

And it's at this level of scale that we start
to see the abnormal patterns of the microvasculature,
which is tangled chaotic development of vessels
with blind ending vessels, shunts loops
and arcades that we don't see in the normal regular pattern
of normal tissues such as here in this renal cortex.

So certainly, at least with animal models
and with superficial structures,
using very high frequency ultrasound imaging,
we are now able to visualize a microvasculature.

We can also perform perfusion imaging.

If we give a high mechanical index pulse
to burst all the bubbles, the bubbles will slowly
perfuse into the imaging space from vessels around,
and we can map regions of interest
and we can plot the arrival of these microbubbles
with low mechanical index imaging
and produce graphs of this type that you'll be familiar
with from dynamic contrast enhanced CT and MRI.

And from these curves, we can derive numbers
that will tell us how well perfused particular
pieces of tissue are.

We can derive perfusion maps
or individual point perfusion profiles.

This is also very useful in cardiology as well as in
imaging of tumors in animal models
and superficial structures.

Now, there are a number of other applications
of contrast enhanced ultrasound
that I don't have time to talk about.

They can be used very effectively in renal imaging
to look at tumors and characterize them
and also distinguish to a certain extent benign
from malignant small renal tumors, at least as effectively
as we can with CT.

But they can also be used very effectively
for vascular imaging.

They can be used to show leaks.

For endovascular aortic repair,
you can detect pseudo aneurysms in pancreatitis
very effectively.

And you can even track acute bleeding from liver and spleen
and abdominal trauma
and show exactly where the bleeding point is.

In fact, I've even treated a few bleeding lesions
by direct injection of thrombin
and fibrin tissue glue, having located the site
of bleeding using contrast enhanced ultrasound.

So just as with CT,
you wouldn't perform an abdominal CT
without the use of contrast.

Certainly we're getting close to the point where
with ultrasound, if we have significant lesions
or significant pathology to investigate contrast,
contrast enhanced ultrasound is equally important
as contrast in CT,

but the research possibilities are continuing to grow.

We've had difficulties until recently
applying ultrasound contrast techniques to organs such
as the prostate, partly
because the high frequencies that we use
for prostate imaging are not really compatible
with the microbubbles that are available.

But different size microbubbles,
different composition microbubbles respond differently
to different frequencies.

And we're gradually learning how to do this.

And we can plot different parameters such as arrival time
of contrast in various points of tissue,
and draw maps of the prostate here, showing
that in this obvious small prostate cancer,
we can see an early arrival time of contrast in the bulk
of the tumor compared to the rest of the prostate.

And this is important when we try to evaluate the effects of different therapies,
medical therapies on prostate tumors such
as anti-angiogenesis agents or anti-androgen therapies.

Here is a simple power Doppler study
of the prostate enhanced by contrast
to give us better visualization
of the degree of vasculature.

After several weeks of anti-androgen therapy, you can see
that the vasculature declines dramatically,
and a switching off of blood flow
through the malignant areas is a characteristic feature
of successful treatment of prostate cancer.

So in Europe, now, contrast
enhanced ultrasound is becoming a fairly standard
technique for the evaluation
of significant abdominal pathology,
but is also forming a bigger
and bigger part of research programs, particularly in
oncology, particularly in the search
for surrogate markers for the immediate effectiveness
of new drug therapies.

Thank you.