Ultrasound Image Optimization - SD

Introduction to Image Optimization

Hi, my name is Tracy Fox from Thomas Jefferson University in Philadelphia, and today we're going to discuss image optimization.

Optimizing the ultrasound image is an important part of the exam.

Every image of every study that we take has to be optimized in order to ensure faithful reproduction of the anatomy and any pathology that's present.

We're going to discuss different techniques, some common and some less common that we can use to make every image we take the best one.

Let's take a look at how we can optimize the image.

We're going to discuss some principles in instrumentation. This is known by its more sinister name, ultrasound physics, but the truth is we have to know how the machine works in order to optimize the image.

We're also going to discuss neology. Neology is the terminology we use to describe what do the knobs and buttons do and how do we get the best image.

By the end of this lecture, you should be able to understand the principles behind the controls on the equipment.

Use these controls in order to optimize the image appropriately and evaluate and identify critical problems within an image and to describe how to correct them.

Knowing Your Machine

One of the most important things about image optimization is to know your machine.

All of the different manufacturers have different ologies, different terminology, and it's important to know the terminology and neology of the machine that you are using every day.

Contact your application specialist.

Have them come to your department and show you different techniques that maybe you didn't know about to optimize the image.

Transducer Selection and Care

Ultrasound is an art and much as an artist uses different brushes in order to create an image.

Ultrasound uses a variety of transducers to produce an ultrasound image.

Choosing the proper transducer, much like choosing the proper brush, is essential to get the best image.

Transducers have different sizes, shapes, and frequencies for different studies.

Remember that transducers are both fragile and quite expensive.

Transducers may range anywhere from $5,000 to $30,000 or more to replace, and many transducers are not covered by a warranty, especially if they consider the damage to be from dropping the transducer.

Make sure you check your manufacturer's transducer manual for a proper care and handling, and most importantly, proper cleaning of the transducer.

Transducers that are damaged with frayed wires or cracked housings need to be replaced immediately.

It's very important to mind the cables and the power cord to be sure that you don't run over them with the machine.

Remember that a good image begins with a well-functioning transducer.

Ultrasound transducers have a special holding spot on the machine.

The manufacturers put these in because they know that this is the best way to store the transducer safely.

Hanging the transducer from the handle is absolutely not recommended.

Part of the problem we run into when we do this is that the heavy head of the transducer may come in contact with the floor and damage the transducer.

Also, the transducer cord wasn't meant to be bent in that way, so please hang your transducers where they're supposed to be.

Also, most of the manufacturers now have hooks to hang the wires and cables of the transducers.

Leaving them on the floor like this invites the chance that the transducer cords will be run over by the machine.

Again, make sure you inspect your transducers for problems with the connector, the wires or the housing of the transducers to ensure that there are no cracks.

Not only will this degrade image quality.

Most importantly, it causes a potential electrical hazard to both you and the patient transducers that are damaged, whether it's the wire that has problems, the transducer cord, or whether there are crystals in the transducer that may be going bad will manifest themselves in your image.

Notice this area of dropout and notice that it comes from the very superficial part of the image and causes a shadow throughout the image.

Again, this is a damaged transducer and should be replaced.

It does degrade image quality

Presets and Basic Image Controls

Manufacturers are very good in giving us presets.

Remember that the presets are a good start, but that should not be where it ends.

You should customize these presets to suit your needs because what the manufacturer thinks is a good image may not meet your definition of a good image and you want it to be pleasing to your eye and to be diagnostically accurate as well.

So use the presets, but be sure to optimize from there.

Let's look at some of the neology of the ultrasound machine.

Freeze Button

Ultrasound machines have a freeze button.

This enables you to freeze the image, evaluate the image for quality, look at the depth, the focal zone, the uniformity of the echoes before you store the image to the packs or other storage system.

When we print the image, meaning when we store the image without freezing it first, we're not evaluating the image for quality before storing it and we run the chance that we may be storing either a blurry or suboptimal image.

So I recommend that you freeze the image before storing it.

Measurements and Calipers

Calipers. There are many things in ultrasound that we're asked to measure.

When you take a measurement with ultrasound, be sure to take your image to store your image to the packs or other storage media with and without calipers.

So you should have two identical images stored, one that has the calipers and an identical image that does not.

This way if the structure needs to be remeasured again, it can be done.

Annotation

When annotating the image, it is very important to properly annotate.

Many of the manufacturers have made it easy for us to annotate with built-in settings.

However, if you're going to label proper labeling is essential.

Notice this image.

Notice that the sonographer told us that this was the right lobe.

It's a sagittal plane and that this mass is located in the mid portion of the gland.

Nobody will ever look at this single image and question where this mass is located.

Proper annotation is essential for the medicolegal implications and complications.

Looking at this image, we see that it's labeled right transverse superior to inferior.

Now where I'd look at a sheet of film or look at these images in series, I would be able to see where this image fit in the series.

However, in the medicolegal situation where this single image may be pulled out of sequence, this image does not stand by itself.

I do not know where this is located.

I don't know if this is in the upper pole, the mid pole or the lower pole, because this image doesn't tell me that it doesn't take long to properly label the image upper, mid, lower lateral, mid medial.

So please properly labeling your images is very important at this part of the study.

Transducer Selection and Frequency Adjustment

When selecting a transducer, it is essential that you choose the proper transducer for the study that you are doing.

Don't be afraid to switch transducers during the study.

You may need a higher frequency transducer or a curvilinear transducer for one part of your study and it may be a lower frequency or maybe a linear transducer to look at the same organ or a different organ in a different way.

Most machines have a button that is either labeled probe or transducer and you are able to select not only the transducer but the type of exam you are performing other machines.

You first select the type of exam that you are performing and from there it will give you a list of transducers to choose from.

All of our transducers today are wide bandwidth or broadband transducers.

This enables us to have multiple frequencies within one transducer.

The frequencies that are available varies by both transducer and manufacturer.

The key point is to always use the highest frequency for the part being imaged.

However, if you're having trouble penetrating an organ, be sure to lower the frequency in order to ensure proper penetration.

Remember, we use the higher frequency transducers for more superficial structures and the lower frequency transducers for the deeper structures.

Let's take a look at this image and evaluate it.

This is an ultrasound of the liver.

Now we notice some possible pathology in the center portion of the liver.

However, notice that the posterior portion of the liver is underpenetrated.

When we evaluate this image, we see this and we should possibly think maybe we need a lower frequency transducer.

This is a phantom with several cysts in it.

This image was taken with an 11 megahertz transducer and this image was taken with the same transducer, but with the frequency change to seven megahertz notice with the 11 megahertz transducer, we are not able to penetrate into the far field.

However, with the seven megahertz transducer, we have much better penetration.

However, remember that while we don't have adequate penetration with the 11 megahertz transducer, we will have higher resolution and in fact, you can see better detail in this structure than in this structure.

Again, always use the highest frequency transducer that you can for the part being imaged.

Focal Zone Adjustment

The focal zone is another knob on the machine that we adjust on every single image that we take.

The focal zone or focal zones should be at or just lower than the area of interest.

Your area of interest may change from image to image.

Therefore, your focal zone may change from image to image.

Don't leave the focal zone in one place during the entire study if you don't have to.

If there's an area of interest in particular that you're looking for, be sure to adjust the focal zone appropriately.

Multiple focal zones or sometimes useful and can give us a better image throughout the depth of field.

However, remember that it does decrease the frame rate, especially for doppler.

It's important to have the focal zone placed in the correct location.

Let's look at these two phantom images.

This caret in this image represents where the focal zone is.

Notice that this focal zone is too high if this is the area of interest, this structure that we're looking at in the middle, these cysts are not well-defined and in fact the borders are ill-defined and we have complete loss of image in the far field with proper focal zone placement.

As you see here, the rest of the image is much better and as you can see, we have improved definition at the region of the focal zone.

Again, when we look at the image such as this liver and we look at the general liver overall, we keep the focal zone at or near the bottom of the image.

So for instance, when I'm evaluating the liver in general, overall I leave the focal zone near the bottom of the image.

However, if I then decide that I wanna evaluate this liver mass, in a better fashion, I need to move the focal zone up and I will get improved definition.

And if we just magnify these images, you'll see with the focal zone placed in a position to evaluate the entirety of the liver, you can see that the definition of this mass is lacking.

The borders are indistinct and we do not have very good contrast resolution.

However, with the focal zone placed in a position to optimize visualization of this lesion, we have improved visualization of the borders and a better image for color doppler and spectral doppler.

Be sure to place your area of interest in the center of the image.

And again, proper focal zone placement is essential.

Overall Gain Adjustment

The overall gain varies in its location from machine to machine.

It may be integrated with the 2D or B mode knob depending on how it's labeled by the manufacturer and sometimes it has other locations as well on the machine.

The overall gain is used to change the brightness of all the dots on the screen uniformly either all the echoes are increased or all the echoes are decreased.

This manufacturer, for example, puts the overall gain knob on a ring that you can turn around the track ball, whereas another manufacturer makes the gain knob stand alone by itself.

Again, if you're not sure how to adjust the overall gain on your machine contact the manufacturer evaluating this image, we can see that all of the echoes in the image are uniformly too dark.

We need to increase the overall gain on this image.

The overall gain on this image is too high.

There's no areas that are particularly too bright or particularly too dark.

The entire image overall is too bright.

In a situation like this, we need to adjust the overall gain time.

Time Gain Compensation (TGC)

Gain compensation or TGC are these gain pods or sliders that you see on your machine.

The TGC is used to correct for the attenuation that occurs as the beam travels deeper.

The goal when adjusting the TGC is to create uniformity of the brightness of the echoes.

Again, we don't want any part of the image at any depth to be too bright or too dark.

So how do we adjust the TGC?

We need to adjust the overall gain to adjust the overall brightness of the image.

The TGC is used to compensate or adjust for the decrease in intensity that occurs with attenuation.

As the beam travels, deeper attenuation increases and the amplitude of the return echo decreases.

Let's look at this example.

I have three reflectors each place deeper than the other.

All of the amplitudes entering the patient will be the same.

So when the first pulse goes in, the sound that comes back will be of a strong amplitude because the beam doesn't have much attenuation occur.

So the return echo will have a fairly high amplitude.

And remember that the stronger the amplitude of the return echo, the brighter the dot you will have on the screen.

The next reflector is deeper.

Therefore more attenuation will occur as the beam travels both to the reflector and back and therefore the return echo will be of a lower amplitude than the more shallow reflector.

Again, the stronger the amplitude of the return echo the brighter the dot.

Because this is a weaker amplitude, the dot will be darker and in fact will be a shade of gray.

The furthest away reflector will have the weakest amplitude because the beam is attenuated both on its way to the reflector and on its way back.

And therefore the return echo for this deep reflector will be a very low amplitude return echo and therefore a very dark dot.

So notice what happens is the beam travels deeper.

We go from brighter dots to darker dots.

This is where TGC adjustment is needed.

Be sure that your TGC does not give you an image where you have an area that is too bright either in the near field to the midfield or the far field.

Be sure to adjust the TGC that so that you have uniformity of the echoes.

Let's look at this image.

Notice this manufacturer shows us where our TGC slope is.

These echoes are all uniformly the same brightness.

This TGC was set properly.

The TGC settings or pods should never be all the way to the left or all the way to the right.

When my gain pods are all the way to the right like this, I have nowhere else to go.

What I recommend is to center the TGC pods, adjust your overall gain and then from there use your TGC to correct for attenuation.

Looking at this phantom, we can see that we're able to see these middle cysts very well.

However, the cysts in the far field are too dark.

Some people would automatically reach for the overall gain in this situation.

However, if we increase the overall gain, we are able to see these far field echoes, however, at least a little better.

However, we have now put echoes into our cysts in the nearer field and these should not have echoes within them.

So the proper way to adjust this image is not with the overall gain, but with the TGC,

Imaging Angle

The imaging angle that we use is important.

Remember that specular reflectors are best seen at 90 degrees.

Therefore, our best angle for 2D imaging is when we're 90 degrees to the reflector, the strongest reflection comes when we're perpendicular.

Let's look at this aorta.

Notice in this aorta, in the distal aorta, I'm able to see a strong aortic wall.

However, when we go more proximally, I lose this aortic wall.

Why did that happen?

What happened to the aortic wall in this location?

Well, the curvilinear transducer, remember, sends out all of its scan lines along the shape of the transducer.

So this scan line coming out this way is 90 degrees to the plane of the aortic wall and therefore this shows up as a bright white line or as a specular reflector.

However, these scan lines that are coming in this direction are not 90 degrees to the aortic wall and therefore do not send a strong reflection back to the transducer.

Depth Adjustment

Using the appropriate depth is very important.

Not only do we not wanna cut off pertinent information such as part of an organ or maybe even pathology, we don't wanna have excess black in the image or wasted space, as I call it at the bottom.

Different manufacturers use different terminology for the depth.

Looking at this longitudinal image of a baby head, notice that we have the occipital bone in the deep field of this image.

When we can see this occipital bone, we know that there's nothing else that we need to see on the other side of that and that we have adequate depth.

However, in this image, I lost the occipital bone.

There may be pathology deep to this image that is being cut off and that I'm not seeing.

Having proper depth is essential to having a good image.

Again, looking at this liver, notice that I don't see the diaphragm in this image.

Therefore it's possible that I'm cutting off pertinent anatomy.

Looking at this image, notice we see the diaphragm.

I know I'm not missing anything.

Zooming and Magnification

Manufacturers also give us the ability to zoom, magnify res.

The image depends on the manufacturer what the terminology is.

Either way, it enables us to enlarge a specific area of interest.

Some machines we must magnify the image while the image is live and on other machines we can magnify the image while the machine is frozen or live.

It depends on your manufacturer.

Making accurate measurements is easier.

When you zoom the image, I always take a reference image.

The reference image gives me the overall picture so that the person interpreting the study can see where I got this image from.

However, then when I wanna measure, in this case the bile duct, I zoom up on the image.

Not only does it enable me to make a more accurate measurement, I'm also able to demonstrate it better for the interpreting physician.

There are two different ways of zooming or magnifying the image.

And again, it depends on your manufacturer what type you have, right?

Zoom redraws the image before it is stored in memory.

In other words, the image must be live for us to zoom this image.

By doing this, by zooming the image before it's stored in memory, we're able to maintain the pixel density, which means we have the best spatial resolution.

Contrast this with read zoom, with read zoom.

The image is already frozen.

Typically, although some machines use read zoom both live and frozen, and in order to zoom the image, it's done by simply enlarging the pixels.

We don't maintain the pixel density when we do this and we end up with a lower spatial resolution image than we would with right zoom.

I always had trouble remembering which one was read zoom and which one was right zoom.

So here's my little cheater, right?

Zoom is the right way to do it.

If you use right zoom, you'll preserve the pixel density and have the best image.

I assure you this is not my house.

But let's look at this picture right here of a house on the New Jersey shore and let's say I wanna zoom this center part of the image.

If I use right zoom, I'm redrawing the image and preserving the pixel density and therefore the image that I have maintains the resolution of the original image.

However, if I zoom this image with read zoom, which would just enlarge the pixels, I do get a larger image.

However, I do not preserve the pixel density and in fact, I can even see the individual pixels within the image.

Acoustic Power Adjustment

All of our machines give us the ability to adjust the acoustic power or output power.

There is what we use and there is what we should use.

What we use with rare exception for most studies is 100% power.

What should we use?

We should use the lowest power necessary to get the best pictures of a structure with adequate penetration.

We have to remember as low as reasonably achievable.

The prudent sonographer or sonologist uses the lowest output power possible and sends the least amount of ultrasound energy into the patient.

Power is very important with respect to patient safety.

The amount of energy we send into the patient is at our fingertips and we have the ability to decrease that amount of power.

We have to remember that bone is a very strong attenuator.

It absorbs the ultrasound and is largely responsible for causing the ultrasound to be converted into heat energy.

Cranial bone in particular is very sensitive to ultrasound heating and in fact the fetal cranial bone and the fetus in general are very susceptible to temperature increases.

So remember when scanning the fetal ultrasound, if you don't need a hundred percent power, the prudent sonographer or sonologist or imager would lower the output power.

Let's look at a phantom, scanned it with a hundred percent power and 25% power and look at the differences.

Notice that on this particular phantom, I still have adequate penetration into the far field.

I'm still able to see the cystic structures that are located in the far field.

However, when we lower the power, we do decrease the ability to penetrate into the far field.

And we also introduce more noise into the image.

You can see clearly that this image is noisier than the image at a hundred percent power.

Either way, remember as low as reasonably achievable.

We're gonna look at some liver images now scan with different powers.

Now I will grant you that this is a relatively thin body habitus patient and not as large as maybe some of our patients are.

But let's look.

This image is scanned with a hundred percent power and this patient was scanned with a power of 50% or a change of minus three db.

Notice that I'm still able to penetrate into the far field.

I still have very good definition of the vasculature of the liver and that there's no noticeable difference between a hundred percent power and 50% power.

Therefore, the prudent sonographer would use the 50% power for this patient at 15% power.

I introduce a little bit more noise into my image, but notice in this patient I'm still adequately able to penetrate into the far field.

Let's take it another notch at 6% power.

Now at this point, we're sending much less sound energy into the patient and therefore the return echoes that come back will also be weak.

However, notice I still have good definition of the hepatic vasculature.

I'm still able to see into the far field.

So between the two choices of using a hundred percent power or using 6% power, what should I do?

Should I increase the power of this image right here?

Or should I maybe just increase the overall gain?

When given the choice between increasing power or increasing overall gain, your answer should always be increase the overall gain.

Remember that gain is a receiver function and that the receiver does has no effect on the amount of power entering the patient tissue.

Harmonic Imaging

Harmonic imaging is another tool at our disposal.

Tissue harmonic imaging gives us a reduction in image artifacts, increases the contrast resolution of the image and provides better visualization of tissue interfaces, especially on our technically difficult patients.

Let's look at that liver mass again, both with fundamental imaging, which is an image without harmonics.

And using tissue harmonic imaging.

Notice that while we are able to see the mass adequately on this patient, the difference in contrast resolution is striking.

Also, we'll notice that the image with the harmonics, that has much more well-defined borders.

Spatial Compounding

Another technique we have is spatial compounding.

Depending on the manufacturer, this will have a variety of names using spatial compounding.

Instead of sending all the scan lines in one direction, we send the scan lines in multiple directions.

This helps eliminate edge shadowing artifact and reduces speckle.

Remember that spatial compounding will decrease the frame rate of your image.

Let's look at these two images of the transverse carotid and notice that this image without spatial compounding, we are able to see a significant amount of speckle within this image.

Notice with spatial compounding, we've drastically reduced the speckle and in fact, we have better definition of some of the soft tissue structures in the neck.

Case Studies in Image Optimization

Let's look at some case studies.

Let's look at some case studies and you decide what is wrong with this image.

Evaluate this image.

Tell me what do you think is wrong with it?

Correct.

The image is underpenetrated.

We're not seeing this far field well at all.

So how would we correct it?

I would increase the TGC in the far field.

I would not adjust the overall gain because I notice that these echoes are adequate Or, It's the far field that's too dark.

If increasing the TGC in the far field does not give us a good image, then we might wanna decrease the frequency.

Instead, let's look at this image and see what we think.

I agree.

There are echoes in the inferior vena cava and there shouldn't be echoes in a normal blood vessel.

When looking at a still image such as this, we're not able to tell if those echoes represent pathology such as clot or that our gain is too high in this image.

I don't know where my focal zone is.

So one of the first things I would check when evaluating this image is where is the focal zone?

Also, you may wanna decrease the farfield TGC to get the echoes out of the IVC.

Now, of course, make sure that those echoes aren't pathology.

So you might wanna put color doppler or some other technique to make sure that they're not real echoes in that IVC.

Let's look at this image.

What do we think?

There are echoes in the aorta.

Also, this image is fairly noisy.

Let's look at the hepatic vein and you see echoes in there.

We should not have echoes in normal blood vessels.

Putting on tissue harmonic imaging will give us better contrast resolution and eliminate those artifactual echoes.

Notice also that we have much better definition of the hepatic vein.

Evaluate this image and tell me what you think.

Notice that the image is labeled sagittal aorta.

So we're being led to believe that it is the aorta we want to evaluate, correct?

The focal zone is here.

Notice that different manufacturers may use different symbols for the focal zone, but with the focal zone placed up at the region of the left lobe of the liver, we do not have adequate visualization of the aorta lower your focal zone to at or below the area of interest.

Notice in this image with the focal zone properly placed, I have much better definition of the far field.

What do you think of this image?

I agree it's way too dark.

For this image where all of the echoes are uniformly too dark, I would increase the overall gain.

This image.

What do you think in this image?

There's either a mass here, which I don't believe, or my near field echoes are too high.

Notice that having improper TGC settings or improper gain settings can actually make something look pathologic.

We need to adjust our TGC in the near field to make all these echoes uniform.

What do you think of this one?

Correct.

In this case, the far field gain is too high and therefore I again would adjust my TGC.

When you hit the freeze button on your image, stop and evaluate.

Look at the image as a whole and ask yourself, what do I think about this image?

So for example, if I hit freeze and I saw this, I might not have noticed while I was scanning that the far field was too bright.

But by hitting freeze and evaluating each and every image and it only takes less than a second, I can be sure that every image I take is my best one.

Now let's look at this structure.

What do we think of this structure?

Correct?

It's not very clear.

The borders are fairly ill-defined.

Well, this image was zoomed using a read zoom on the machine.

However, if we use write zoom notice I have much better definition because we're able to redraw the image, preserve the pixel density, and maintain spatial resolution.

What do we think of this image?

I agree.

The depth is not properly set.

I don't see the diaphragm in this image, so therefore I am cutting off part of this image and I may be missing pathology.

Therefore, we need to adjust the image depth.

Remember that a lot of our pathology, especially in the liver, occurs on the edges of the liver.

And I've seen masses missed because the depth was not properly set.

What do we think of this one?

I think the focal zones in the perfect spot.

I think the echoes are uniform and adequate and I can see the diaphragm.

So I'm done right, wrong.

The depth in this image is not properly set because I have too much wasted space.

If you want to optimize your image, you need to have the largest image that you can have on the screen without cutting off anything.

For one thing.

It makes the image easier to read and it's easier on the eyes and it's the optimal way to do it.

So again, for this image, properly adjust your image depth.

Now let's look at this structure.

I think there's a few issues with this image.

Yeah, I agree.

This image is way too dark.

We can barely see any definition in this liver whatsoever.

And also remember what I said about measuring something.

If you're gonna measure a small structure such as this two and a half millimeter bile duct, you really want to zoom up on it before measuring.

So I would, on this image, I would adjust my gain.

I would take my overall picture just to show the reading or interpreting physician that this is a structure I'm gonna be measuring next.

And then the image that I take will be zoomed.

And I will take the zoomed image both with and without calipers.

See how much easier it is to see exactly what I'm measuring.

And let's look at this image and what do we think of this?

Now before you say it's a bad transducer, or maybe somebody ran over the cable of this transducer, notice that this dropout, these echoes don't occur at the very surface of the image.

The problem with this image is that you don't have enough gel, especially when scanning the groin region.

It's very easy to get air underneath the transducer because of the anatomy in that region.

So for this study, I would add more gel.

Adding gel will eliminate the air that is trapped in between the transducer and the patient's skin.

However, let's look more closely at this transducer.

Notice that in the very corner, I do have an area of dropout.

When I add more gel, if this is caused by air, this will disappear, but this won't.

And let's take a look, right.

Once I added more gel, I had better visualization because I don't have any air in there.

But notice that this area of dropout persists throughout the study.

That's because this transducer is damaged.

Conclusion

So remember, you are the artist and the patient is your canvas.

Be sure to properly take every image as optimally as possible, evaluate your images and be sure that you're using the best image you can to demonstrate anatomy both normal and abnormal.

And you'll take the best pictures that you are able.

Thank you very much.