0:00

Briefly, DBT is a form of limited angle tomography, in which multiple low

0:04

dose exposures are obtained in arc like projection, and then reconstructed

0:08

to form 1 millimeter slices through the breast, which mimics like a CT

0:12

exam. Of course, acquisition and reconstruction algorithms vary by vendor.

0:18

You can see in this table, this paper looked at four primary vendors.

0:22

You can see the vendors and systems change or are different in terms

0:27

of their scan angle, the overall scan time, the number of projections that

0:31

they get, whether the tube motion is continuous, meaning that the arc

0:36

is continuously moving while the machine is acquiring images, or it's a

0:41

step and shoot kind of movement. Detector and materials are largely the

0:46

same. Some of the reconstruction algorithms are different.

0:50

And then the reconstruction resolution is slightly different amongst the

0:53

different systems. That being said, this was, the paper was put together

0:57

in 2019 and companies are constantly improving the reconstruction algorithms

1:03

and detector resolution. So even some of this has changed since that paper

1:08

was published. And of course that's a constantly moving sort of thing.

1:14

So with DBT, there are three primary datasets that are obtained.

1:20

First, the raw projection images, which you saw in that video just there.

1:24

These are reconstructed into 1 millimeter slices. And then

1:28

some practices acquire a separate full field digital mammogram, a 2D mammogram

1:35

in so called combo modes. That would be DBT plus a 2D mammogram. Some practices

1:42

use what's called a synthesized mammogram. I'll talk about that in just

1:45

a second. So the way this works in general is that we have,

1:51

the breast is positioned the same way it is for 2D mammography. A

1:55

detector on the bottom, compression paddle, breast in between.

1:59

The difference here of course is the X ray tube is now moving

2:03

in this arc like projection. And in this graphic here, you can see

2:06

that we have our tube moving at different angles and taking separate projection

2:10

images of the breast. And in our system, that's 15 projections.

2:15

These are reconstructed into individual Tomo slices, which we can then scroll

2:20

through or move through to view them. And then other systems at some

2:26

point, like I mentioned, will reconstruct those slices or those raw projection

2:31

images into a more flat file, which mimics a 2D mammogram in this so

2:35

called synthesized 2D mammogram or SM view. As you can imagine,

2:42

doing combination mode, DBT plus full field digital mammography confers

2:46

a higher radiation dose as compared to 2D alone. It's approximately double.

2:51

You have two separate image acquisitions. To address this, in 2013,

2:57

the FDA approved the use of these synthesized mammograms created from the

3:01

DBT dataset. Now, importantly, that approval was dependent on

3:07

looking at the DBT dataset and the synthesized dataset. So it's not just

3:12

one or the other. So meaning that you can't, for example,

3:14

if you, let's say, read a study from

3:16

outside facility where they send you just the synthesized mammogram, that

3:19

is not adequate for interpretation. Utilizing SM makes the total radiation

3:25

dose similar to full field digital mammography. It allows easier comparison

3:30

to prior exams. And studies have shown that it's not inferior to combination

3:34

mode DBT and there's no significant difference in calcifications. That being

3:39

said, this particular point is still sort of hotly debated out there in

3:42

the literature. There are some practices that prefer the use of combination

3:48

mode, particularly for the evaluation of calcifications because the resolution

3:52

is much better. And there are some practices that are willing to accept

3:56

a decrease in resolution to improve the overall radiation dose. Importantly,

4:01

the combination mode, 2D plus DBT, is still within the appropriate dose

4:05

limits. So it's okay to do.