Automated Ultrasound for Breast Screening and Whole Breast Surveys - SD

Introduction

I am Ellen Mendelson, professor of radiology at Northwestern in Chicago.

Today I'll be talking about automated ultrasound for breast cancer screening and the use of automated ultrasound for whole breast surveys.

In today's presentation, I'd like to talk about the need for supplemental breast cancer screening, the results of the Akron 6, 6, 6, 6 trial.

What methods we might use for screening, either handheld ultrasound, automated ultrasound, and why one may be preferable to the other.

I'll provide for you a bit of historical perspective and show you how an automated scanner can function in the very busy breast center workflow.

We'll also talk a little bit about a multicenter trial of lesion detection with handheld versus automated scanning that will proceed widespread screening.

What is Screening?

What is screening?

Screening is the detection of unsuspected disease and the disease should be of high prevalence as breast cancer.

The screening method should be affordable, repeatable, accessible of low risk, and provide effective treatment.

And effective treatment should be provided.

Current Accepted Methods for Breast Cancer Screening

What is accepted for breast cancer screening at the current time, it's mammography and physical examination.

And for the highest risk group, contrast enhanced breast MRI.

What about a role for ultrasound?

Breast Cancer Statistics

Breast cancer is the number one cancer for women with 180,000 to over 200,000 new cases per year in the US for 2010.

The expected number of cases is 207,000 invasive cases and about 53,000, situ. Two cases.

Breast cancer accounts for about 40,000 deaths per year with mortality rates reducing, by about 2.3% per year since 1990 with earlier detection and better treatment as explanations for the reduction.

Breast Cancer Screening Recommendations in the United States

What are the breast cancer screening recommendations in the United States?

Mammography annually starting at age 40 for most women, and their mortality reduction of over 15% from randomized control trials provides the evidence.

There has been a publication, of the US Preventive Services Task Force, a government advisory panel, based on statistical models to start mammography, not at 40 in the us but at age 50, and to perform it for women of average risk, not high risk, but of average risk every other year.

Why this new recommendation?

Could it have an economic basis in this, time of healthcare cost crisis?

We have found many younger cancer patients, whose ages range between 35 and 49.

And should we start screening at age 50, we would expect, many more late stage cancer diagnoses.

MRI, is part of the recommendations for screening, and there it's for highest risk women annually in addition to mammography, particularly where the lifetime risk is greater than 20% ultrasound for dense breasts and elevated breast cancer risk.

But up to now, there has been no recommendation from the American Cancer Society or other bodies for screening with ultrasound.

We expect a change this year or next year, in the American College of Radiology practice guideline for the performance of breast ultrasound in 2007.

The most recent version ultrasound is listed, as an area of research for screening in 2011.

We would like to see a change, in that we recognize data supporting ultrasound screening in addition to mammography, not instead of for women with dense breasts at high risk of breast cancer.

Limitations of Mammography

What are the limitations of mammography?

These are something we must recognize there. In here.

Mammography does a fantastic job in breast cancer screening, but it has limitations as every modality does.

Contrast limitations are what provides the problems in, mammography.

We see ma can we see masses in dense breast tissue?

I'll show you an example, of what is really white on white, and it makes a mass hard to see in the dense white fibro ular tissue and this type of tissue, a normal type of tissue component in breast can reduce sensitivity of mammography by up to 50%.

Any woman, young or old, can have dense breast tissue.

The women who are young, younger than 40, younger than 30 have denser breasts, and very young women have higher radio sensitivity.

Just for your orientation, I've labeled on these medial lateral OB OBL views back to back of the left and right breast.

The muscle. This triangular area that you see, in the upper part of the image, the fibro tissue is dense and white.

The skin you can trace, here's the nipple and the dark areas of tissue are fat in the breast.

Now, on the left side, because this is a mirror image, the left breast has a pseudo cancer that I've placed against the fat.

It's an asterisk, and there's a second one here.

Can you see it? I could ask you if it's in the other breast or somewhere in the same.

It happens to be in the same breast and it's in the upper portion.

And with contrast now where I've shown it in black, it's easily, perceived.

So white on white is difficult, and that is the contrast limitation you'll find in mammography.

Risk Factors for Breast Cancer

Who is at risk for breast cancer.

Women whose age is increasing women with a family history of breast cancer and first degree relatives, and those are mothers, sisters and daughters and fathers, especially, when the first degree relative is premenopausal and has bilateral breast cancer.

When there has been a previous biopsy, with a histology of lobular carcinoma in situ two, atypical lobular hyperplasia, atypical ductal hyperplasia, and the last I've listed in this group, ductal carcinoma in situ two is not really a high risk marker.

We count as cancer, personal history of breast cancer and radiation for Hodgkin's lymphoma, where one might expect, the risk of cancer, breast cancer, to increase after as soon as eight or nine years have passed.

Hereditary breast cancer accounts for about 10% or up to 10% of the breast cancers we find.

And there, if you think of 200,000 as the total in the US per year, we might say between 10 and 20,000 cancers are of the hereditary type BRCA one and two are abnormal breast cancer genes.

And the women who have those have very high risk for BRCA one.

It's 87% lifetime risk, and 50% of those women, have their disease found by age 50.

ACRIN 6666 Trial

A very important study completed now, with the first year results having been published in 2008 is the American College of Radiology Imaging Network.

The Akron study, 6 6, 6 6, which is performed, was performed in conjunction with the National Cancer Institute.

And what happened in this study was that handheld ultrasound physician performed by experienced ultra sonologist, in 2,800 women who had dense breasts and had high risk, were in interpreted independently from mammography, also performed for these women.

Final assessment categories were assigned one through five based on birads, the American College of Radiology, breast Imaging Reporting and data analysis, systems, a feature anana analytic approach for ultrasound, for mammography.

And for MRI, the goals of the study were to find the independent contribution of ultrasound and to do a cost analysis.

The cost analysis is still being, worked on and has not yet been published.

The first year results of Akron 6 6 6 6 were published in JAMA in 2008.

And there the diag diagnostic accuracy of mammography plus ultrasound was found to be 91% with mammography alone, 78%.

The diagnostic yield of mammography alone was 7.6 for a thousand women screened, which is how you calculate it.

And when ultrasound was added, it rose to 11.8 per thousand.

Women's screen. Of these cancers that were found with ultrasound 12 were found only with ultrasound, not with mammography.

11 of them were invasive.

And of the women whose axillary status we knew, eight of nine were had negative lymph nodes, the supplemental cancer yield of ultrasound, then we found in a range of 1.1, which is low, to 7.2 per thousand women screen with an average of 4.2, per thousand women screen.

There were a large percentage of false positives, and we were concerned about that.

Only 8.9% of the biopsies, showed cancer.

But we recognized that the first year was a prevalence screen for ultrasound.

In other words, there had been no previous ultrasound studies to compare, to compare with as we had for the mammograms.

Many of the women had years of mammographic, studies that we could compare with, with the comparison studies available, we expect that years two and three, will be better in terms of positive predictive value.

Nevertheless, lack of ultrasound specificity has been something of concern for decades.

Limitations of Ultrasound as a Modality

As we look at ultrasound as a modality, we always think of it as operator dependent.

And some of the past limitations of ultrasound worry, it's inadequate spatial resolution.

It's small field of view, five centimeter squared, really maximally.

These over overlying speckle, coarse noisy image, poor inconsistent scanning technique.

And for breast inexperience with ultrasound anatomy and pathology, its diagnostic use was limited to cyst versus solid, without permission really to look at solid masses and try to distinguish among them.

And there was no standard for lesion analysis and reporting, before B rads was published in 2003 with automated scanning rather than a small field of view, the entire breast anatomy is depicted no matter what the method of automation is, whether it's supine, whether it's prone water, bath, or helical in terms of its acquisition, you see the entire breast, and these acquisitions are volumetric, so they enable reconstructions of other views just as we can with CT and with mr.

So that if, as we concentrate today on the supine, automated scanner, the acquisition is a linear acquisition in horizontal plane, and it's a very wide up to 15 centimeter acquisition.

And this acquisition will allow a reconstructed coronal plane.

The entire breast and coronal plane is new for ultrasound, and we are learning its applications.

There is also a vertical reconstruction.

So we have three planes as we do, say with stereotactically guided biopsies.

An X plane, which is horizontal, a y plane, which is vertical, and a Z plane, which is anterior posterior.

It's the depth plane.

There is no radiation, no ionizing radiation from ultrasound, showing you how we can depict anatomy, more completely.

This is anatomy deep in the breast near the chest wall.

And the, at the top of the image, you see ribs down deep in the image is the thorax.

Above you see ribs in periodic array above the ribs, you see muscle, and the detail is good enough for you to visualize the muscle bundles that comprise the petal muscle.

And above that, some fat.

There's very little breast tissue in this one image as you can see.

And then at the top, is the skin.

Little yellow square is showing you where on the breast, anteriorly, is the nipple.

But go to the coronal reconstruction.

And what you see is in detail is the anatomy of the sternum, with aco osteochondral junctions and ribs emanating from it, and all the portions of the sternum are visible.

And here in this vertical reconstruction, the periodicity of the ribs is something that you can appreciate.

There is a diagrammatic representation of your location of scanning.

You see that on the bottom of the image here.

If you look at the anatomy of the breast in full coronal view, again, reconstruction at the skin, you'll see pores.

And I wanted to show you this image because this patient has had a breast reduction.

And the scar, that vertical scar from the real region down to the inframammary fold is visible.

The symmetry is easily seen, and the architecture is readily appreciable, when it's normal and when it's disturbed here in another patient, is the anatomy, at the skin.

The nipple is recognizable.

You can see a portion of the arreola.

And what you also see are prominent pores in her skin.

This patient has inflammatory breast cancer, late stage breast cancer, and this was at the skin of the affected breast.

And it looks very different from the two cases I showed you previously.

The skin is thick. It has edema within it.

The MR study is shown for correlation here, the very thick skin.

It's a centimeter thick, where normally it should be two millimeters in thickness.

And these lines that you see going through the coronal view are represent edema.

And the swelling of the skin and tissues is also appreciable here, around the nipple and in the periphery of the breast.

This slice is reconstructed a little bit more deeply in the breast, at the level of one of the many masses, that were found within her breast.

It was grade three invasive ductal carcinoma, that had affected the lymphatics.

Examples of Automated Scanning

Another case, and I'll show you several cases, it's probably the easiest way to appreciate, what automated scanning can do for you.

Here's a 47-year-old with no family history of breast cancer and a mass that was tender and palpable behind her right nipple.

And it was somewhere hidden in this white fibro ular tissue.

Targeted ultrasound handheld initially showed right at the site of the palpable mass, a two centimeter simple cyst at 12 o'clock.

And here it is, no other abnormalities.

And we then went to an automated scan.

Many women have multiple masses, both cystic and solid, and we wanted to see if perhaps, this patient had other areas of involvement.

the.is on her nipple, and what you see adjacent to it, and this is deeper than the skin level, is another area.

It's hypoechoic.

You can see granular tissue looking white, and the margins of this area are somewhat irregular.

We went then this is the acquisition scan, and you can see here, with a portion of the skin not visualized, on this scan, that there is a cystic and a solid component of this tubular structure, which was a duct that contained and here in the vertical reconstruction, an area irregular, and within the glandular tissue, of this breast, a solid component and anterior to it tic component.

So we have here evidence of an abnormality that requires further evaluation.

We did a biopsy, and this was an intraductal papilloma unsuspected.

This is muscle.

And here the shadowing that you see is artifactual, where the breast tissue and the fat lobules, which are hypoechoic, intersect.

One thing that the automated scan enables you to do because of its long or its wide acquisition, is to measure, the distance between two lesions.

This makes, it easier when you are looking at cancers, to assess multifocality or possibly multi centricity.

Here we have a distance that isn't very far.

You have centimeter markers, and this was about a centimeter and a half, between the cyst and the papilloma.

And here on, just for correlation is the handheld depiction of the same lesion, here with wide field of view here, with the small field of view.

And I think the detectability and the diagnostic, potential, of the automated scanner can be seen.

Developments in Handheld Scanning

As we look at breast ultrasound, concomitant with development of automated scans, we've seen developments in handheld scanning.

And how has breast handheld scanning changed in recent years?

Many manufacturers are offering elasticity assessment, tissue stiffness assessment, and there are various methods, that they employ, to enable the stiffness assessment to be done.

Matrix and volume acquisitions, or multiplanar enabled multiplanar reconstructions, as we've seen with automated scans, but also available with handheld small fields of view, transducers at high resolution.

And all of these volume acquisitions, will enable a cad, a computer aided detection, or diagnosis based on birads to be developed.

And there has been development in this area.

3D and four D now is available at high resolution.

There are calcification algorithms that some, systems manufacturers are developing.

But again, we are working in small fields of view, and I think we still have need for automation for the wild, wider fields of view.

Specificity and Sensitivity in Ultrasound

Looking at ultrasound, looking at any modality, we have to look at specificity and sensitivity.

Specificity is defined as the true negatives, in the numerator over a denominator comprising false positives plus true negatives.

And what it really means is the probability of normal findings when no cancer exists.

Sensitivity, true positives over true positives, plus false negatives.

And there we're talking about false negatives that are largely unknown without a link to a tumor registry.

They only come to light later, sometimes in sad situations.

So just some examples of the specificity problem I'm showing you here.

Five sonograms and one little mammogram.

And of these, can you determine which one of these five is benign?

I'll tell you, but it would be very difficult, for you to guess.

It's this, this happens to be a duct, with a very thick inflammatory wall.

And this was benign.

This is the open lumen of the duct, in cross section.

This is a grade three invasive ductal cancer.

This is a grade three invasive ductal cancer looking quite different.

The first one I showed you, if the technique were, were less than optimal, might resemble a cyst and an error could be made.

And here a very subtle change in the architecture of the granular tissue, which is light gray, and the fat above it, which is darker gray, a small invasive lobular carcinoma.

And here also looking quite circumscribed, but really isn't, is a mass behind the nipple that was also a papillary carcinoma.

So again, just just to show you how difficult it can be, elastography, can show you tissue stiffness, a hypothesis being that cancers are stiff and benign.

Things are soft, which is not always the case.

Some cancers are soft, many mucinous cancers are soft.

And there are benign very firm tissues as in stromal fibrosis.

But in general, this hypothesis is helpful.

And here, a cancer, in its b mode image on the left and on the right, a color overlay, that shows you the stiffest reddest portions of the lesion.

Usually the scale is seen at the right side, of the image.

And here you can see hard and soft.

In the United States, the FDA has not yet approved a quantitative measure of stiffness.

No matter what the method of elastography, we are awaiting the FDA's change of heart.

I think that will be helpful.

The specificity problem here, you need clinical correlation.

Both of these are images are from the same patient, and here is architectural distortion.

You can see that the ular tissue, this is muscle below, and fat above, but the granular tissue is distorted, and they're all of these little hypoechoic areas in it.

Two months later, this is what this patient's area of concern looked like.

Her thi skin was then thickened.

There was edema, around it.

And this is a postpartum mastitis that has progressed to abscess formation two months later.

So you can't be sure you'd need the clinical history.

You need correlation, of all of the imaging studies if others have been done.

Screening with Handheld Instrumentation

Can we screen with handheld instrumentation?

This is a question that we really have to ask ourselves.

And we look to the Akron study for some guidance in physician performed, handheld bilateral ultrasound.

It took 19 minutes on average, for a study to be completed.

And 19 minutes of physician time is that possible in the United States in 2010, in 2011, in 2012.

As we do more and more of these studies, it's uncompensated, and there's really no way that this can be allowed.

But can a sonographer or a technologist perform, and the breast imager, a physician interpret a sonographer or a technologist, is really not licensed to, to render an interpretation.

And we don't have data. It's an unanswered question.

If this kind of screening, could be done, it's possible that large scale, instruction, and experience might give a positive answer there, but we haven't gotten to that point yet.

Survey scanning with handheld small field of view is really a mosaic made of small many pieces that are knitted together by, by the brain.

And as one scans, one is interpreting at the same time, you just move over the breast and knit together these little mosaics.

So with handheld ultrasound, the performer and the interpreter in real time are the same person, with limited physician experience and time, and decreased availability of this study.

I think the demand for handheld ultrasound performed by a physician cannot be met.

So there's a need for something else.

And the something else is probably automation.

Requirements for Automation

And when we look to automation, what might we want?

Do we want multimodality fusion?

When automation was first considered, many of the mammography interpreters, had little experience with ultrasound, and they sought ultrasound fusion with mammography, mr As the learning curve now, increases, and many mammography and ultrasound, breast imagers, are learning.

MRI, there's some thought about fusion with ultrasound and MRI particularly, for biopsy, purposes.

Now, what about patient positioning?

Supine prone upright, one principle of ultrasound is the minimization of tissue thickness, so that you have higher resolution, and the penetration, depth is lessened.

Can we do prone automation?

We can do it. And, we can see some of the results there, that are fairly good.

But supine in terms of the theory and physics of ultrasound would be the way to go.

3D multiplanar reconstruction is important, and volumetric acquisitions can also support CAD systems for detection.

The wide fields of view, 14 to 15 centimeters, in 2D in the supine automated scanners, really counters the small field of view compromise, of handheld what might be applications of automated systems.

We can look at and four multiple masses, to examine, to be examined at time one, and later followed up.

And you'll be able to identify, mass one as mass one and MAs two as MAs two, and know which ones are which ones, which has been a criticism of follow up of lesions, mass lesions, with handheld ultrasound.

So we can use it also for follow up of benign and probably benign lesions.

Birads characterization and structured reporting, can be entered into these systems and in newly diagnosed cancers or patients suspected, of having cancer to search for additional abnormalities.

In other words, extensive disease evaluations can be performed with automated systems, just as they can with MRI, which at the current time is the most sensitive modality, for these studies.

And also then for screening, historically automated ultrasound is not new.

Historical Perspective on Automated Ultrasound

Ultrasound itself, dates back to 1951, with a mode ultrasound.

And the potential of x-ray mammography began to be realized in the 1950s with a work of Gershen Cone Egan, John Wolf, and many others.

And at that time, interest in ultrasound diminished, there was fear of ionizing radiation, which led to a, renewal of interest in ultrasound in the 1960s and seventies, when there was development of ultrasound systems, whole breast systems of in Japan, Australia, and Germany.

And then in the 1980s, in the us we had whole breast systems developed by Technica and Lab Sonics, the latter, a, a company and located in Indianapolis that no longer exists.

The design and construction of a breast scanner began in Australia in 1963.

And if you look at some of these scans, or the table and the water bath there, that provided, some compression on the breast, just as we see now.

It's little bit, more primitive looking, but ultimately, and ours is more streamlined, but the resemblance, I think, is there at that time.

He also had compound scanning, but it was from more than one transducer.

And here we can see the intersection of beams, which provides you with a spatial compounding just as we have, in our current high resolution transducers.

So here's an example of the lab sonics.

This is more recent than 1963.

The transducer would come down and traverse, this, plastic shell, in which there was water, and then making contact with the breast, a polyethylene bag.

Here's the breast and the type of image that was generated at that time.

This is a fibroadenoma, but each of the fat lobules is definable, and you can see down to the chest wall.

And, the skin is nicely depicted in both of its layers.

Current Supine Automated Scanner

So, looking into the future, we have an automated scanner supine, a linear high resolution scanner, with wide field of view.

And we started working with this scanner, a year or so ago.

And it acquires in three planes, orthogonal, coronal, and it has multiplane or reconstructions.

As we've discussed.

The image is generated through a mechanical traversal of a transducer from inferior to superior.

The contact with the breast made by a mesh screen in and a thick lotion placed, placed on the patient's breast, to reduce any possibility of air.

There's mild compression to immobilize the breast.

And the transducer frequency that we are using currently, is a wide aperture broadband with, transducer 14 of, of frequency range, 14 down to five.

There's a handheld, system, upon which the automated system, is integrated.

And the views per breasted for complete coverage are two to three, currently three and AP view, a lateral view and a medial view.

With all of the supine scanners, there is limited visualization of axillary tissue in the axilla.

Should you want to scan it, is ideally suited to handheld in interrogation.

It, the non-physician acquisition time for two breaths.

In our study done last year, was 10 minutes, 60 seconds per view, and electronic image review, was then performed.

And there are many user defined hanging protocols, which I'll show you.

And how you use these hanging protocols has a large impact on the efficiency of your review and interpretation.

The recall rates of screen patients, was a workflow and cost concern, and, Dr. Tula dis in Rochester looked informally at, 1,300 patients, whom she screened and had a low recall rate of 4%.

This was reported, but at the R-S-N-A-N-A couple of years ago.

But the results were not published.

This is an example of this automated system, and there are others.

This isn't the only one.

There's a second supine system as well that is FDA approved and on the market, but, and very similar, in its design to this.

Here, the transducer, is housed, and the mesh screen is boarded by this gray zone.

It comes down against the patient's breast.

You can see it here, as it makes contact.

Once the positioning is assured, a lock is engaged and the transducer automatically goes from the lower breast where it's placed, to the upper breast and then releases, and then this would be the AP view, and then the other views, can be positioned for, subsequently and additional scans obtained.

So this is the right anterior posterior view, and we center the nipple, with the probe barrow located in the center of the footprint.

As you see here, the right lateral view, is done with the patient not moving, but the breast moved towards the medial aspect, exposing the lateral portion of the breast, to the transducer.

So here, is how that would be, done.

And the portion of the breast that would be depicted is shown in blue.

Similarly, again, without moving the patient, the breast tissue for the medial view is moved laterally.

And the area that you see is on the, the area of the image is shown in blue.

Pathology in Automated Views

What can you see? And how does pathology show itself, in these views?

Here is the acquisition view.

It's located in the upper, outer quadrant of this breast.

The nipple in this portion of the coronal image in the left lateral view, is for your orientation and then going to the upper outer quadrant.

The lateral aspect, as you can see here, is an area where spicules extend from a hypo coic mass, possibly calcification seen within it.

But on the acquisition view, which correlates with it, and where the crosshairs intersect is, are the three correlative views, the acquisition view, the coronal reconstruction, and then the vertical reconstruction.

What you do see is the mass with the architectural distortion around it, the tissue planes are being destroyed by the advancing cancer, and the skin is thickened above the cancer and being pulled down.

This is skin retraction.

You can see it also in the view, that is vertical and its extension towards the chest wall will not tell you, it doesn't tell you on ultrasound ever whether or not the chest wall, is invaded by tumor.

For that, you need enhancement on MRI, to be able to be confident, in saying that you have chest wall invasion, but the coronal view, is remarkable in showing you the architectural distortion, face on.

And there you can see the rest of the architecture better on an AP view here, but the focus is on this cancer.

Do you see this other little hole? This was a cyst.

And frequently, you will be able to pick out lesions easily, in the coronal view.

One way to look at these images is in this tomographic depiction, and you can select the slice sickness, as you go anteriorly post through the posterior aspect, of the lesion.

Other ways to look.

And there are many tools that you can see here in this toolbar, which enable you to look any way you would like.

And it's important to find a particular way that you use each time, to look at cases, to shorten the period of interpretation, so that you can be efficient, in your reporting.

This is an example of a single slice hanging protocol, where the acquisition, view and the vertical view are smaller than the reconstructed coronal view.

And this is a right lateral view.

Accentuating the lateral tissue of the right breast.

Orient yourself to this scan, with the nipple seen in the yellow box and the magnifier over three cysts.

One is a small simple cyst, another is just a simple cyst.

And the third, and you can see on both the acquisition view and the coronal view, a small calcification in the dependent portion, of the cyst.

So the resolution is quite good here.

Looking outside of the magnifier, you can also see additional, small cysts and some dilated ducts.

The correlation among the three images, is excellent, and your overall orientation, can be shown in this round diagram.

And in the diagram of the two breasts depth, you can always tell by the scale that you'll see at the side of the acquisition and the vertical image.

And there is also a depth measure, for the coronal view as well.

So 3D orientation is possible.

One thing important if one is going to use these scanners for follow up is the ability to look side by side at last year's or the year before's image compared with the current one.

And these, the workstations enable, the comparisons to be shown.

Another example of how you might see a cancer, on three different views and its relationship to the nipple, again, in the yellow box, and here on the coronal view, it's easy to pick out this irregular mass.

And you can also evaluate, the rest of the architecture as well, and perhaps find a second area resolution is, is good on this acquisition, with stickles extending from the hypoechoic mass and calcification seen centrally.

And, within the mass itself in the coronal view, you can see the mass itself here, the coronal view is quite small, and you get an overview, and the location, the of the scan is about a centimeter and a half from the skin, what the vertical view can show you, and you can zoom any of these views also, to see them larger if you need to, is the periodicity of some of the artifacts that you would expect from the intersection of curved surfaces.

The fat lobules with the glandular tissue beneath it.

And here you, it will help you, to exclude, the possibility of small masses that cause posterior acoustic shadowing.

So these can help you, in many instances.

Another case showing you two cancers, they are both high grade and one here a second here.

And the distance between them is measurable.

Another case here, 47-year-old with an area that was palpable in the right upper outer quadrant.

And this may be another application, for automated ultrasound because of its wide field of view and its depiction of larger areas of the breast.

So on the mammogram, you have heterogeneously dense or scattered fibro densities with an area of asymmetry in the upper right breast compared with a corresponding area of the upper left breast, which is fattier.

And going to the other view of a screening mammogram, or the cranial coddle view.

In this case, the mammogram wasn't for screening.

The patient had a palpable mass, so this was a diagnostic study and the marker shows where she felt something.

And it's this rounded area that is in the same area, of the asymmetry on the medial lateral oblique view.

And again, you can see in the outer left breast compared with the right, that the area is fattier and there's less glandular tissue.

So the asymmetry holes in two views, and one would wonder about an underlying mass to explain it.

Well, the first thing that we did was to do a handheld sonogram, and it was done initially by the sonographer and then checked by the physician.

And in talking with the patient and with, in palpating that area, it became evident that this area, that the patient felt, was more apparent when she was in a seated position.

So the handheld ultrasound was done in that position.

What was found in the handheld ultrasound was a rounded area that of echogenic fibro tissue, similar in size and in the location of the palpable area in question.

And this area of glandular tissue came almost up to the skin, and it was flanked by hypoechoic fat lobules here on the right and on the left, we saw down to the muscle.

So we knew we were looking at the entire breast at that location.

Well, could that explain it?

How confident were we that, that the ultrasound, showed no abnormality, in fact, explained the area as an area of granular tissue, and we looked at the other side and it didn't look like this, area of glandular tissue that rose to the skin was firmer than the fat loles that flanked it.

Well, here the automated skin really did help, in increasing our confidence.

And we were able to tell the patient that there was no abnormality in this area of breast tissue that she felt, and we were able to explain the reasons, for her feeling it, here in the coronal view, no change in the architecture.

It was normal.

The cross hairs, are centered in the region, of the palpable finding.

I've taken them off in the other view so that you could see the area better, as just glandular tissue, and asymmetric clump of it, rising to the skin when there are sufficient data and enough studies are done, to support this as an indication for ultrasound and to rely on it, to exclude any abnormality, I think we will, be able to include this as an indication, for automated ultrasound in a multicenter study.

Upcoming Multicenter Trial

We are looking at breast lesion detection at six sites, three in the us, two in Germany, and one in Japan.

That will, involve 250 patients per site for a total of 1500 patients to see if the automated scanner, the supine automated scanner, has detection capabilities that are equivalent to or better than those of handheld physician performed examinations.

We expect that they will be, and that ultimately the method for screening in the future, will be with automation.

Conclusion

In conclusion, automated scanners are important breast imaging modalities.

Once supplemental ultrasound screening is improved, supported by the Akron 6, 6, 6, 6 trial data, we'll find increasing utilization of automated scanners.

The automated multicenter trial, we hope will establish, that the Akron 6, 6, 6 6 handheld results are achievable or better with automated scans, and that the automated scans will also, in addition to detection capabilities, will have, diagnostic potential as well.

And in the future, we expect that automated survey ultrasound will be the method, used for screening, because of its reproducibility in follow up, its detection and capabilities and follow up potential for benign masses, it's rapid performance without involving a physician for performing it.

It will allow handheld characterization and ultrasound guided biopsy, and we feel that it can be integrated efficiently into the breast imaging workflow.

Thank you.