Volume Sonography - A New Era in Efficiency and Diagnosis in Obstetrics - SD

Introduction

My name is Beryl Raf, and I'm a professor of radiology
and obstetrics and gynecology at Harvard Medical School.
One of my passions lately is 3D ultrasound volume
sonography, as I call it.
And I think it's a very, very important new
modality in ultrasound
and imaging, that I think has a huge future.
And I'm gonna be talking today about the applications
of 3D volume sonography in obstetrics.

Volume sonography, otherwise known as 3D
ultrasound is a new era in efficiency
and diagnosis in fetal imaging,
which I personally am very, very excited about.
And I think for a while, 3D ultrasound got the wrong
impression out there
because people thought
that it was all about just getting pretty pictures
of the face.
And in fact, it's a lot more than getting pretty pictures.
3D is volume scanning
and it's all about how we can display that volume.
With 3D ultrasound, you acquire an entire volume,
which then permits you to render an infinite number
of slices within the volume
and of types of displays of the information.
You can either slice it and dice it,
or you can render the entire volume
as a surface, for example.
But the volume contains all of the information
that's available and it's up to us
to display the information that's within the volume
to advantage so that we can show what we wanna show.
So we've only just begun to explore the ways
that we can display the information within that volume.
So today I just want to talk about really four
of the main display modes that we have
to tease out the information,
the sonographic information within the volume.

Surface Rendering

The first one we'll talk about is surface
rendering at an interface.
And for that you need to have enough fluid at the interface
to be able to render a surface.
The second one we'll talk about is
the three orthogonal planes.
These are three planes that are at right angles
to each other that intersect at one point,
and it's the main way in which the volumes come
up when you first take them.
And you can navigate through
that volume using those three planes.
The inverse mode we'll talk about in a moment,
and the tomographic cuts
are parallel cuts.
They're parallel to each other, much like a CT and an MR.

So let's start out with surface rendering,
which is entered interface,
and that shows an image of the surface of an object.
Of course, there has to be enough fluid in front
of the object to form an interface
that can be rendered.
And this display can be accomplished
in lots of different modes.
You can have a soft tissue mode,
you can have a skeletal mode, you can dial through a lot
of post-processing modes to display what you want.
For example, this is a seven
and a half week fetus surface rendered,
showing the little head,
which is bent over and the yolk sack.
And this is the three orthogonal planes showing you,
where that little tiny fetus is
and showing you the green line is going
to give us the interface where we're gonna render.
And then in the lower right hand corner,
you see the rendered image showing you the little head
and the yolk sack
underneath the fetus is looking like
it's sitting on the yolk sack.
And then of course you have the little limb buds
that can be visualized
and compare that to the drawing of what
a fetus looks like at a crown rump of 10 millimeters.

Now here we have at eight and a half to nine weeks.
You can already see the enhancement that you get
by doing a 3D surface rendering compared
to a standard 2D crown rump image.
You can see the fetal ear, you can see the little arm
and leg buds much better than you can in 2D.
And you already get an idea of what this
embryo looks like.
And you can compare that to the drawing, showing
that the ear is
actually fairly low at this stage, which is normal.
And the ear eventually rises up
to be about at the level of the orbit.

At 12 weeks, we get a lot more
information, surface information about the fetus.
You can see the digits, you can see the eye, you can see a lot
of details better than you can see them in 2D.
These are two little guys that came in
to have their nuchal translucency measured.
And here we're imaging them from the back.
And I just wanna show that you can see all four ears.
You can see both ears in both fetuses showing you
that in fact this is a great time to image the ears,
by doing surface rendering.
And it's not something that we normally think
of doing at 12 weeks is to take a look at the ears.

This is a fetus that has amniotic band syndrome.
This fetus was about 11 weeks.
And you can see that there are a lot of abnormalities here.
There is a large encephalocele here.
There is a large anterior abdominal wall
defect right here.
You can actually see the bands right in here and in there.
But surface rendering shows you the information even better.
It actually shows you a severe cleft right in the
center of the face right here.
And you can see that that cleft is a destructive cleft in
that it comes all the way into the middle of the face
and goes up into the large encephalocele
or a craniorachischisis if you will, where you have a large amount
of brain that is not covered by skull.
And you can see anteriorly also the very large anterior
abdominal wall defect in this fetus
with severe amniotic band syndrome.

Now sometimes it's important to try
and get more information about a dead fetus.
And many times, unfortunately,
we are called upon to image a patient who's bleeding.
And in fact, when we look, we find
that the heart is no longer beating
and the fetus has passed away.
And of course, invariably the mother will say,
what happened?
Is this gonna happen to me again? What did I do?
And so forth. And if you can give her more information as
to perhaps the fact
that this fetus was abnormal, it is helpful.
And in this case, we imaged the fetus
in a surface rendering mode
and found that kink in the spine right here.
That ended up just not being visualized at all
on the 2D image.
So this was very helpful to the mother to know
that there was probably some sort of abnormality
that was quite severe.

Now here's another case in the late second trimester
where we're doing a standard survey
and we're looking for the bladder.
And where is the bladder?
Well, sometimes it empties out
as it does on a regular basis,
but then it starts to fill very quickly.
And if you wait for a little bit,
you will see the bladder fill.
Now here we are in the lower aspect of the fetus.
This is the lower end of the spine here.
And then we have the insertion of the
umbilical cord right here and the genitals,
but we don't have a bladder.
Now the genitals do look somewhat abnormal
and I'm not really sure if we can tell whether it's a boy
or a girl here, but there is no bladder.
Well, one can infer
that there is bladder exstrophy in this case
because the bladder not being filled is probably lying on the
anterior abdominal wall.
But what better to show that than a 3D ultrasound
and here on the surface
of the anterior abdominal wall is in fact the bladder,
which is splayed open in a bladder exstrophy right here.
And this was in fact very helpful.

Here's another case that came in
because of a kink in the spine
seen on another scan at another facility.
And here is the kink right here.
It's a hemivertebra, it's a vertebral abnormality.
And when I took the entire volume
to look at the entire spine all in one image,
I was able to render it as a skeletal mode.
And lo and behold, I saw that there was a rib missing.
The absence of this rib was not something
that I appreciated in 2D
and I went back to my 2D to see if I could appreciate it.
And sure enough, by working hard, I was able
to see it in 2D.
But this was identified first in 3D
and it was not something that I had appreciated originally.

Another thing that 3D surface rendering is very helpful
for is to get the global picture of
how the fetus is lying.
2D enables you only to get one thin slice at a time.
And so you depend on getting a lot of little thin slices
and conjuring up in your mind
what the fetus looks like in three dimensions.
But here you can actually display
what the fetus looks like in three dimensions
and get an entire view
of the whole uterus all at once.
And in this case, this fetus who has arthrogryposis
has some contractures of the upper
and the lower extremities.
You can see that the fetus has hyperextension at the
knee and club feet.
You can see that better on this side
and you can also see the clubbed hand
or the clenched hand because of the contractures
and this fetus is not moving at all.

Now 3D has helped us also look at the palate.
Now in this case, as you know,
looking at the lip is not that hard to do.
We can look at cleft lip
and palate quite easily with both 2D and with 3D.
But looking at the soft palate is very difficult
and really has not been reliably done with
regular ultrasound.
But in this case, if you take a volume of a fetal profile,
and of course the display is always in the upper left hand
corner and here I'm displaying the profile
but I flipped it upside down.
So the chin is up here and the forehead is down there
and you bring up your render box so that all
that you're going to render is actually in this
little box right here.
And your point of view is this green line
and you put the point of view inside the mouth
and look basically atlas at the palate.
And what do you see? You actually see
rendered the entire palate.
And this is a great new method
of looking at the palate inside the mouth.

Now let's look at an abnormal one here again,
the fetus has been imaged in profile
and then flipped upside down with the chin here
and the forehead there, I've put in my render box
with my point of view, green box right in the mouth.
And we're looking up at the palate.
And what do we have here?
We have a rent
or a space right in the middle
of the palate, a linear space.
And in fact, this fetus had a normal face other than some
micrognathia or a small chin, but the lip was normal.
So what do we have here? We have a cleft palate.
You can see the cleft right here
in the middle of the palate.
And you can compare that to the normal palate right here,
which is all in one piece.
Compare that to the path specimen of this fetus
who had multiple congenital abnormalities.

So let's look at some clefts.
Now in this case we're looking at a cleft lip and palate.
And those are easy to identify both in 2D and in 3D.
But the 3D is helpful for the parents to appreciate
what the child is going to look like.
And also for the plastic surgeon
who is not particularly trained in looking at
cross-sectional imaging and ultrasound.
So here you can see
how similar the cleft is looking to the newborn because there's a large hole right here
and some collapse of one side of the nose.

Here's another example. This is a cleft lip only.
Now here the palate is intact.
I'm not showing you the palate here, but
the palate is intact
and you can see here that there's a cleft.
This is on one side and this one has it on both sides here.
And this is particularly helpful
because it really shows what the child is gonna look like
and we who are trained in ultrasound can scan in 2D
and sort of go in and out of the face
and conjure up in our mind what that cleft looks like.
But it's very hard to convey that information to the parents
and to the plastic surgeon who is gonna try
and explain to the parents what kind
of surgery is going to be needed.
So this really makes it real and makes it helpful.

And here is another fetus
that had a bilateral cleft lip and palate.
And just to show you how well these can be fixed
by the plastic surgeon, this is the child
after it had the surgical repair really a superb outcome.

Now let's look at some ears
and look at some dysmorphology of ears.
You can see in this fetus with Down syndrome
that the ear is in fact very small.
And here is the ear right here and here.
And it is quite small for the size of this fetal head.
Now we don't have many criteria for
what ears should look like,
and 3D is gonna give us the opportunity to look at
what the ear should look like
and what the criteria should be for a normal sized ear.
Here is another fetus with Down syndrome.
And in this case the ear is probably not necessarily small,
but it's very low set and rotated forward.
The bottom of the ear is rotated forward,
which is also quite abnormal.
And the dysmorphologists have a lot of information
and experience in diagnosing
low set ears and in knowing what that means
and what the syndromes are that can be associated
with low set ears.
And now we have that capability
because we're able to render the face
and look at the dysmorphology that the
geneticists do after birth.

Here is this fetus
with Down syndrome showing their protruding tongue
because fetuses with Down syndrome have poor tone
and the tongue tends to protrude.

Another dysmorphic fetus.
This fetus has a neural tube defect.
And I'm sure you guessed right
that the fetus has a neural tube defect.
Even though I'm not showing you a picture of the spine,
what I am showing you is a picture of the head
with a pointy head known as the lemon sign.
This looks like a lemon. And I can also show you
that the cerebellum is wrapped around the brainstem
and it looks like a banana.
And that's the banana sign, the lemon sign
and the banana sign are associated
with the Arnold-Chiari type two malformation,
which is in turn associated
with an opening in the spine somewhere along the spine.
So what this tells me is this fetus has an opening somewhere
along the spine, but we don't know really
what these fetuses really look like.
And now with the 3D surface rendering,
we can take a look at the face of this fetus.
You can see the pointy forehead.
Very funny looking ear here.
A small ear, probably somewhat low set
'cause the ear normally comes up
to about the level of the orbit.
And it looks like there's also a small chin micrognathia.

This fetus has a funny ear as well,
but it also has hypertelorism.
The eyes are too close together and the chin is very small.
These are dysmorphic changes that we used
to have a hard time discerning with 2D ultrasound,
but now with 3D we can display them almost as well
as they are displayed later when the baby is born.

And we're looking, here's another application
for surface rendering and that is the genitals.
Now here are two different fetuses with ambiguous genitalia.
This one here on the left has a phallus looking
area right there.
And I don't know whether these are labia
or this is part of a scrotum.
And again, the same thing, a phallus
and then either labia or scrotum.
What do we do now? Well, we try
to figure out if this is a boy
or a girl by doing an amniocentesis.
Well, this fetus is a girl XX and this one is a boy.
Well, they still look the same to me.
So then we do our 3D rendering,
which is going to be very useful.
And you can see here that there is masculinization
of the clitoris right here
and that this is a labia right on either side
and that's very different compared to what we see here,
which is clearly a boy that has severe hypospadias
where the penis is actually pulled down below the scrotum.
That's why it's called penile scrotal transposition.
It's a severe form of hypospadias.

Now look at how incredibly different they look in 3D,
but how similar they look in 2D.
And this has been very helpful because then we can ask our
colleagues in urology to help us make these diagnoses
because now we can render them in a way
that they're used to looking at them.

Three Orthogonal Planes

All right, let's move on to another type of display.
The three right angled planes,
the multiplanar reconstruction.
This is a type of display that shows you three planes
that are at right angles to each other.
And it enables us to manipulate the planes
to get just the right slice
and that slice can actually be completely reconstructed.
This type of imaging gives us the opportunity
to make the acquisition plane irrelevant
and to be able to display any slice at all that we want.

This is a profile.
Now this fetus was in a
supine position when this image was taken,
when this volume was taken.
And yet I can display here the coronal view
of this face even though I could not scan
in that orientation.

Now here's another fetal head here,
which shows you that we can scan in this view.
That's the acquisition view, showing you the
lateral ventricles here.
We can also scan right at right angles to it.
And in this case we're seeing the corpus callosum,
but this view, which is the axial view, very much like a CT
or an MR, is a view that we can't scan in.
We can only reconstruct it.
We cannot in this particular fetus actually come in
and get this picture directly with 2D.

How is that useful? Well, it is incredibly useful
because in this third trimester fetus whose head was way
down, I wanted to image the corpus callosum,
but there was no way in either of the two planes
that I could scan in
that I could display the corpus callosum.
So I took a sweep right here
and reconstructed the midline
to show the corpus callosum.
So I can reconstruct this view because it is in the volume
and all I have to do is figure out how
to display the one view that I need within one whole volume.

Here's another fetus that also has agenesis
of the corpus callosum.
And this little dot right here is actually on a cyst.
Now this is the intersection between all three planes.
So this dot represents the same space as that
and the same space as this.
The dot is on the cyst.
This is an interhemispheric cyst associated with agenesis
of the corpus callosum.
And in this reconstructed view here you can see the parallel
orientation of the lateral ventricles, which is typical
of agenesis of the corpus callosum.
And that is in the reconstructed view that we cannot provide
unless we take a volume and reconstruct it.

And this is one last example.
And in this case we're in the skeletal mode.
As you know, we have lots of different modes
that we can display and here we're showing the skeleton,
much like the case that I showed you
where the rib was missing.
And this enables us
to display any view within the volume of the spine.

Inverse Mode

Now let's go on to the inverse mode.
The inverse mode is very interesting.
What it is is it's a rendering mode
of the entire volume,
but it takes everything that is cystic within that volume
and makes a cast out of it so that everything
that cystic becomes opaque
and everything that is solid melts away.
So that what we're able to see, instead
of in this case having one slice showing you these different
cysts in a pelvic kidney,
you take everything that's cystic and make it opaque
and you can see that this is a hydronephrosis.

Now this hydronephrosis comes in and out of multiple planes.
So there is not one plane that will help you display this,
but if you take the entire volume
and push the one button to get the inverse mode,
then everything that cystic will become white.

Here is another hydronephrosis.
And when we push the button, not only can you see
that the hydronephrosis is right here,
but you can see the fibrillated ends
of the hydronephrosis.
And you can also see the follicle
that was sitting in the ovary that was
above it in a different plane.

Now getting back also hydronephrosis is particularly
helpful to show in
the inverse mode
because these hydronephroses can go in
and out of many planes
and there is no plane here even in any reconstructed plane
that will show you the whole hydronephrosis.
But getting back to the fetus, now this is a fetus
that has hydrocephalus
and it also has a funny cystic area right here in the
frontal lobe
and the frontal portion of the lateral ventricle.

Now let's display that with inverse mode. And here we are.
This is a cast of all of the fluid areas within the brain
and you can see the lateral ventricles,
the entire lateral ventricles as well
as those cystic components in the frontal area.

Here is a fetus that has hydronephrosis
and you can see the fluid in the kidney in the collecting
system of the kidney and in the ureter as well.
But this is only one slice
and this is the entire kidney,
the entire ureter all the way down
to the bladder seeing in inverse mode.

And one last example, this is a fetus that has
urinary obstruction, it has posterior urethral valves
and you can see the very, very large bladder
that this fetus has.
You can see the posterior urethra.
And in this case actually you can also display the kidneys
and you can see that our dot,
you see our little dot right here,
our dot right there is on the descending aorta.
And so that's what all
of these three images have in common is this one dot.
So in fact what these three images show is
the bladder and the dilated kidneys in this patient
with posterior urethral valves.

Now you push one button and you have the inverse mode
and this shows you now the bladder
and the posterior urethra very nicely.

Tomographic Ultrasound Imaging (TUI)

Going on to the last type of imaging technique that I want
to talk about, and this is actually one
of the most important ones is the tomographic ultrasound
imaging, TUI.
And it's multiple slices that are all parallel
to each other, very much like a CT or an MR.
You can set the distance between the slices to be one, two
or three or four millimeters apart depending on
your subject or what you want to do.
But basically here is a fetus that has mild hydronephrosis
and here I've taken one volume
and in this volume I'm displaying these cuts going from the
front of the head to the back
and the display goes from the front all the way
through the cavum septum all the way through to the back
to the choroids and back here to the occipital horns.

Now having seen that, then I can say, okay, I'm gonna go to
a different orientation, I'm gonna go to image B.
And here we are now going from side to side
and I've arranged my slices to go from side to side.
Now this is not a new volume, this is the same volume.
I'm just slicing it in a different orientation.
And here I can go from one side with a choroid
and the lateral ventricles clear
through the midline showing the corpus callosum
and all the way out the other side to the other ventricle.

Now there's still one more orientation, the C plane.
And here is that coronal plane.
This is the one that you can't even scan
and this is the one that you have to reconstruct completely.
I'm going from top to bottom here, very much like a CT scan
and you can see the axial view of the entire
brain this way.

Now you can also do this in a normal case.
Now here for example, is a normal fetal head
and what can we get from this normal fetal head?
Take one volume and what do we have?
We can look from the top to bottom of this volume.
We can see here that we've got the posterior fossa.
We've actually can measure the nuchal fold back here.
Got the cisterna magna back here.
We can actually grab our BPD either here or here.
Here we've got the cavum septum.
We're coming up into the lateral ventricles and the choroids
and we can actually see the walls of lateral ventricle
and go all the way up to the top.
We've actually got everything we need here
to archive the anatomy of this fetal head.
We just take one volume and we're done.

Here's another volume of the body of the fetus
and this shows the chest and abdomen
and we're going side to side with these slices.
And what do we have here?
Here we have the diaphragm, we have the right lung.
We can actually see the gallbladder right here.
We can see the aortic outflow tract right there.
You can also see the pulmonary artery outflow tract.
You can see the insertion of the cord.
And then you can see the stomach
as we come into the left side way over to the left side.
And you can also see the insertion of the cord here
and perhaps maybe even a little piece
of the bladder there you can see a lot of information.
You can also see the spine here in the back.
So there is enormous benefit to this type of display.
It's a type of display that CT and MR has had for years.
All the relevant anatomy can be captured all at once
and the patient can be on the table
for a vastly reduced period of time.
And then you can do the entire scan on a virtual
scan on a review station
because you have all of the information
as though you still had the patient.

So that's why I think that 3D volume imaging is one
of the most important advances in sonography
because finally, ultrasound has acquired the same
capabilities and hopefully the same type of displays
as CT and MR.
Because when you think about it, the technique
of taking one 2D image at a time,
which is a very thin slice while standing next
to the patient for over 20 minutes
is really no longer feasible.
2D really is not enough anymore.
We need to have an entire volume that we can navigate
through and reconstruct any plane that we want within
that acquisition and gain control over all of the planes
and all of the anatomy and the displays.

Study on Efficiency of 3D Ultrasound

Now I was a radiology resident a long, long time ago
and at that time
a CT scan took 30 minutes to do.
Now, recently I had a CT scan
and it only took two minutes to do.
And why hasn't ultrasound kept up with this?
Why are we still taking 30 minutes to do our studies?
So I did a study to look at whether we could put the fetus
through a 3D ultrasound machine, much like I was put
through the CT scanner to get my CT scan.
And can we generate a large amount of data of the anatomy,
a huge volume that includes all the anatomy
and that can be read elsewhere offsite.

So for 50 consecutive fetuses undergoing their structural
surveys between 17
and 21 weeks, we did their standard 2D scan, one
of our eight sonographers did.
And then we further took five additional volumes
that were designed to encompass all of the fetal anatomy.
These included the fetal head, thorax, abdomen, face,
and lower extremities.
And I'll show you an example of some of these two
of these volumes in a moment.
And these were designed to try
and capture all of the fetal anatomy so
that the patient could go home
and all the fetal anatomy could be retained.

Now a couple of weeks
after the original scans,
the volumes were then reviewed on a separate workstation
and we looked at the time it took to obtain the volumes,
the time it took to actually do the surveys offline.
We timed ourselves
to do the surveys offline on the five volumes including
measuring the BPD and the femur.
We looked at the completeness of the survey.
Did we were we able to identify all of the landmarks
and we took a look and compared the time it took to do all
of that to the time it took the sonographer
to do the standard 2D scan,
there were three physicians involved,
myself and my two partners.
And we reviewed the downloaded volumes.
We timed ourselves to see how long it took
to review those volumes offline.
And of course the physicians were blinded to any
of the original 2D information and findings.

Now here's an example of a volume taken
to evaluate the heart and let's look and see.
Now I took this volume
and I'm gonna enlarge plane A, this is plane A
and I'm gonna magnify it so I can see.
Now I'm gonna navigate up and down through this volume.
Now I'm navigating up and down.
I'm getting a four chamber view of the heart.
Here's my four chamber view
and I'm gonna go up a little bit above the heart
and I'm gonna identify the pulmonary artery.
There is the pulmonary artery. I'm gonna move my dot.
My dot is my pivot point.
I'm gonna move my dot onto the pulmonary artery
and I'm gonna pivot using a Y rotation.
I'm gonna open up that pulmonary artery
so I can see it lengthwise.
Okay, now I'm gonna put my dot on the aorta, that's a piece
of the aorta just behind the pulmonary artery.
I'm gonna open it up again using my Y rotation.
And here is my aortic outflow.

So now I've examined the heart within the volume
that was taken by somebody else who may not know the anatomy
of the heart, but I had the volume and I could examine it.
Now I'm gonna rotate the volume back so
that it's in transverse section,
which is the way I like to scan.
And I'm gonna go up and down
and now I'm seeing the stomach when it come down
through the kidneys, the anterior abdominal wall.
And I'm gonna go down and see the
bladder, there's the bladder.
Okay, now I'm gonna move my dot
onto the spine and now I'm gonna pivot.
And this is my new pivot point.
So now I'm pivoting on the spine
and I can examine the spine from top to bottom very nicely.
And I can then navigate through the volume in
that orientation, that new orientation,
and show all sorts of different planes.
Look at the diaphragm, look at the bladder again,
look at the bowel and look at the spine.

Now as you can imagine there is overlap of these volumes
and you can see the five volumes are listed here.
And the head volume will often
identify or the head,
but it'll often identify also the arms that tend
to be in front of the head.
Here for example, is the abdominal volume.
The abdominal volume right here will identify most
of the abdominal type findings,
but it'll also identify the heart about half the time.
So there is some overlap
with these volumes that is very helpful.

And here's another example. This is a face volume.
This was taken to evaluate the face,
but we can actually evaluate the brain.
And let me show you how we do that.
I'm gonna take my A plane, okay?
And I'm gonna move my, I'm gonna enlarge it.
First I'm gonna move my dot into the center of the face
and I'm gonna use that to pivot with my Y axis.
I'm gonna move my dot again.
So now we're in the middle of the brain
and I'm gonna orient myself so that I can see the brain.
There's the posterior fossa
and now I'm in a situation
where I can actually measure the BPD.
I'm gonna take my measuring tools.
So here's my measuring tool, my 2D distance,
and I've measured my BPD.
I'm gonna go back to the main menu
and now I'm going to navigate up and down.
I've got the posterior fossa there,
I've got the lateral ventricles, cavum septum, choroids.
I've got all the anatomy that I want here in the brain.
And now I'm gonna move my dot onto the spine
and I'm gonna orient myself at right angles
to the initial orientation and I can navigate back
and forth seeing the spine
and all the way out to the face,
which I'll show you in a moment.
And there is the face, the nose, and the lips.

So we can see a lot of anatomy within that volume.
So more than 90% of all of the anatomic landmarks
that we look for were identified successfully on this
virtual scan as compared to the 2D scans.
And all the three physicians were equally successful
at reconstructing the images
and identifying all of the findings.

Now the time was really quite astounding
because the mean time that it we needed
to obtain the five volumes was under two minutes,
two minutes of table time.
The mean time needed to do the reconstructions varied
between 4.8 and 5.5 minutes among the three physicians.
So let's just say five minutes.
So the mean time for the entire volume scan was between six
and a half and seven minutes.
So you can see if you compare the time needed
to do the 2D survey, which was 19 minutes,
that there was a big difference.
In fact, the volume scan saved 12 to 13 minutes per patient
and that's room time.
That's time in the room on the machine
for our 50 patients.
The total time saved by using this 3D technique
was over 10 hours.
This suggests an incredible opportunity
to increase the efficiency of our scanning in the future
and also to standardize it so
that we are not as operator dependent.

Implications and Future of 3D Ultrasound

Now just to finish up with a few issues that may have come
to mind as I went through this.
How is the patient gonna perceive this?
Is the patient gonna want more scanning
time to view the baby?
What happens when the scan is incomplete?
How about the sonographer's point of view?
Are they going to be able to keep their jobs
and have we increased the time that the physician must put in
As far as the patient perception,
this method is probably going to be mostly used
to archive the images that we need.
Instead of taking them one image at a time,
we can take a whole sweep, archive all of the images
with one button and probably still have plenty of time
to do a brief realtime component for the patient.
And also to look at the heart, the cervix and the placenta.
Of course, the patient will need to be educated as to
what comprises the medical part of the ultrasound.
No other modality prolongs the exam beyond
what is medically indicated
because it's very expensive
to spend time on the ultrasound machine.

About five to 10% of the exams may require additional views.
It's mostly of the hands and the feet.
That may not be totally included within these volumes in
some fetuses, but that should be apparent
before the patient leaves.
As we can review these volumes very quickly
before the patient leaves in mammography centers,
the screening is often done without the physician on site
and there is a small number of patients that get a callback
for additional views.
And of course this type
of scanning can be done as a screening type.
I'm not advocating this for high risk patients.

The impact on sonographers I think is going
to be very important.
The images are gonna be less operator dependent.
Sonographers will spend less time scanning
and this may improve some of their repetitive injuries,
problems that they have
and that they deal with long term.
And then the sonographer can be called upon
to reconstruct the views offline
and that will vary their task
and hopefully reduce their injuries.

And the impact on the physician well will have some
beautiful displays very similar to CT and MR
because ultrasound is now able to provide some
of these large volumes that CT and MR enjoys.
And really ultrasound as we know it today is not going
to be economically feasible long term.
A 30 to 40 minute session
to do a full structural survey, one-on-one
with a skilled person cannot be reimbursable
or cost effective for long.
And plus, once you have these volumes,
the physician can then reconstruct any plane even if it
didn't happen to be a plane that the sonographer
had planned to image.
So that much more information is available within the
volume than the information
that was actually photographed by the sonographer.

So can you imagine if CT scans still took 30 minutes?
We need to move ultrasound into the better efficiency,
which was previously inaccessible to ultrasound,
but is now available to CT and MR.

So what do we have in the future?
Well, I think 3D ultrasound is likely
to really change completely the way ultrasound is done
and certainly the way ultrasound is archived so
that we don't have to take one little slice at a time,
but we can take volumes and archive the slices that we want.
Improvements in the reconstructed image quality, the speed
and the real time is gonna be available in all
planes with full resolution.
That will be even better than we have it today.
We will have new innovative displays such
as huge volumes that'll be all inclusive,
can even be displayed in real time perhaps,
and manipulated completely offline, very similar to CT
and MR, but perhaps completely in real time,
which will be better than CT
and perhaps 2D only scanning
instruments will be relegated
to the tiny little miniaturized scanners
that you can hold in your pocket.

So lots of changes in the future.
Volume scanning is only in its infancy
and I think it's a very exciting time
to be in the ultrasound community now.
Thank you very much.