0:00

Let's take a look at a 3 Tesla MRI using a

0:05

pelvic array coil and compare it to the 1.5T

0:07

comp vignette, which you'll

0:10

see labeled on MRI online.

0:13

This is an 81-year-old, and of course, as you

0:16

get older, the incidence of low-grade cancers

0:18

goes up, so it's really important to discriminate

0:21

who has an aggressive cancer and who doesn't.

0:24

And you're going to do that by size, by boundary

0:27

crossing, by hypervascularity, But especially by the

0:31

diffusion characteristics and the morphology of the

0:35

lesion, whether it's crossing key boundaries or not.

0:38

This patient had a negative biopsy,

0:41

uh, two years prior to this study.

0:43

And the PSA wasn't really elevated.

0:45

It was hovering about, around 2.4.

0:47

18 00:00:48,169 --> 00:00:50,380 But yet the clinician felt

0:50

something on physical examination.

0:52

And this is the perfect patient to have an MR.

0:56

You know, the clinical exam's

0:58

positive, the PSA, non-contributory.

1:01

And the patient is 81 years of age.

1:05

The patient also had a family history.

1:07

So let's look at the axial T2-weighted image.

1:10

And there is an area of ill-defined

1:13

hypointensity in the PZL and PZP, or

1:17

posterolateral and posterior aspect of the gland.

1:21

And it's ill-defined with a

1:23

fair amount of capsular contact.

1:26

We lose to some degree the nice black line

1:29

as it comes around, so we're, we're a little

1:30

concerned about microcapsular invasion.

1:34

We still do see the little dots of the neurovascular

1:38

bundle in the 5 o'clock at 7 o'clock position,

1:42

5 o'clock over here, 7 o'clock over here.

1:45

And so we turn to our T1-weighted image.

1:48

I'd prefer to have an axial T1, but I'm blessed

1:53

at least with this coronal, and I'm looking at

1:55

the neurovascular bundle right in the area of

1:57

the tumor, so I'm scrolling backwards to that

1:59

area of ill-defined character, and I really don't

2:02

see any effacement of the fat on either side.

2:05

I still see the veins and the neurovascular

2:08

structures, that's a nerve right there.

2:10

And so they're unencumbered by the tumor.

2:13

So we're not going to give this patient gross invasion.

2:17

There may be microcapsular invasion.

2:20

Now let's turn our attention to the

2:22

diffusion image, by the way, at 3 Tesla.

2:25

And I'd like you to compare back and

2:26

forth this 3 Tesla study with the 1.

2:30

5 comp image that was a PI 5, like this

2:34

one, to look at the Disparity and image

2:37

quality of which there really isn't any.

2:40

And here is your axial diffusion

2:43

image at a higher B value.

2:45

Here's the low B value, here's the higher B value.

2:47

Not all that bright for this tumor.

2:49

Let's go to another image at another locus.

2:53

Not all that bright.

2:54

I'm going to window it a little tighter.

2:56

Let's go back to the one that I think

2:57

is a little more impressive right there.

2:59

So not as bright as I might like.

3:02

Let's go to something with a slightly

3:04

higher B value, a single image.

3:07

And it's a little bit brighter.

3:09

So, why on this image with the low B

3:12

value and the high B value, low B value,

3:14

high B value, why wasn't it as bright?

3:17

Well, the main reason was that the B value

3:21

for this particular image wasn't high enough.

3:24

I want the B value to be 1200 or greater.

3:27

But let's talk about, for a moment, the causes

3:31

for false positive and false negative diffusion

3:35

imaging because diffusion imaging is the

3:38

major criteria for peripheral zone lesions.

3:41

That's point one.

3:44

Point two, causes of false positive and negative

3:48

diffusion imaging, which we'll get to in a second.

3:51

And point three, you have to know

3:54

what causes diffusion restriction.

3:56

And what does?

3:58

Hypercellularity, cell packing, necrosis, as

4:03

many of us see in the brain, say with heroin

4:06

encephalopathy, infarcts, like we see in the

4:10

brain, with destruction or breakdown of the cell

4:13

membrane where the water gets trapped in the cell,

4:17

hyperviscosity, as might occur with, say, a cerebral

4:21

abscess or an epidermoid, and inciting desmoplasia,

4:27

where the interstitium becomes very stiff and firm.

4:31

So now let's talk about item number two of

4:35

our three points, which is the causes for

4:38

false positive and negative diffusion.

4:41

One obvious cause, the spatial

4:43

resolution just isn't good enough.

4:46

Typically, you're not going to get great

4:48

diffusion unless you've got a lesion that's at

4:50

least five millimeters in cubic centimeters.

4:56

Mucinous tumors, fortunately they're

4:58

rare, but they don't restrict.

5:01

But nine very cellular nodules In the TZ, and

5:05

sometimes proliferating into the PZ, these can

5:10

restrict, so this is a cause of a false positive.

5:15

So in the central region, or TZ of the prostate,

5:18

diffusion restriction is not a major criteria.

5:22

It's a supplement.

5:24

If it's positive, it helps with

5:26

a suspicious morphologic lesion.

5:29

If it's negative, it helps.

5:31

It really doesn't help you all that much.

5:33

But you can't use diffusion imaging on

5:36

its own in the TZ or transitional zone.

5:39

Abscesses.

5:41

These are going to restrict inside

5:44

where there's hyperviscosity.

5:46

This is also known as the vanilla donut sign.

5:50

Because you've got the white vanilla in the

5:52

middle on the diffusion-weighted image and the

5:56

solid portion of the donut around the outside.

6:00

In a tumor, the diffusion restriction

6:03

is more peripherally oriented.

6:05

So the brighter signal is going to be on the

6:07

outside, the darker signal on the diffusion.

6:10

Not the ADC map, on the diffusion.

6:13

The brighter signal will be on the outside,

6:15

the darker signal will be on the inside,

6:17

so we call this the chocolate donut sign.

6:22

And then, as we said, any cytotoxic or

6:25

necrotic edema can diffusion restrict, and

6:28

this can rarely occur in the prostate gland.

6:32

Now, some other critical causes of,

6:34

of failure to diffusion restrict.

6:37

Well, one is low-grade tumor, you know, Gleason 5 or 6.

6:41

The B value, as we said, not high enough.

6:43

Gotta have that B value 1,200 or more.

6:46

And then another one, perhaps the reason why this

6:50

B value did not at 1,000 show the tumor as well.

6:54

One, the B value wasn't high enough,

6:56

and two, there's air in the rectum.

6:59

So air in the rectum may inhibit your ability

7:02

to get a nice, crisp diffusion restriction.

7:06

So let's draw these donut signs.

7:08

So in abscess, we're gonna have lower signal intensity.

7:13

around the outside on the DWI, say

7:16

B1600, and on the inside where there's

7:19

hyperviscosity, it's going to be bright.

7:23

That's our vanilla donut sign.

7:26

If we've got a tumor, it's going to be the opposite.

7:31

We're going to have diffusion

7:33

restriction on the outside.

7:35

We're probably going to have less

7:37

diffusion restriction on the inside.

7:41

The so-called chocolate doughnut sign.

7:43

And if you want to be a purist, maybe

7:45

I should make the chocolate brown.

7:47

Because chocolate is brown.

7:49

So this would be the tumor chocolate doughnut sign.

7:53

And in the middle, this would be

7:56

the abscess vanilla doughnut sign.

8:00

Now I want to share one other thing

8:01

with you on this 3 Tesla study.

8:03

So let's get out of our drawing tool for

8:06

a minute, and let's go to the DCE MRI.

8:11

The Dynamic Contrast-Enhanced MRI.

8:14

So let's match it up with our Axial T2.

8:17

Let's go right to that spot, and here's our mask.

8:22

Now immediately, right after the contrast arrives.

8:25

Mask, contrast arrival, boom.

8:28

Yes, these nodules in the middle

8:30

are enhancing, very common.

8:33

So, you really can't use Dynamic Contrast-Enhanced MRI

8:37

unless you have a big charcoal-like smearing mass with

8:41

very intense enhancement and very obvious washout.

8:46

In the periphery, we do see, as an isolated

8:48

phenomenon, first 10 seconds, enhancement,

8:53

In the PZL, lateral, and PZP, posterior.

8:58

Let's keep going.

8:58

It's getting brighter.

9:01

It's getting brighter.

9:03

It's getting brighter.

9:05

And now it's getting a little less

9:07

bright compared to the rest of the gland.

9:10

So it's starting to wash out.

9:13

So the curve in this one is

9:15

looking something like this.

9:17

Came up pretty quick.

9:20

And it takes a little while and then

9:22

it starts to drop off towards the end.

9:26

Now you can quantify this with a CAD, with a color map.

9:30

Let's see what the color map said about it.

9:31

Let's see if it picked it up.

9:34

And indeed it did.

9:35

Right there.

9:36

It also picked up these innumerable TZ

9:40

central region of the gland areas of

9:41

hypervascularity which are extremely common.

9:45

So our DCE MRI said that the lesion was hypervascular,

9:50

fire engine red, and that had a pretty decent washout.

9:58

So, I don't rely on the hue very

10:00

much in CAD for the prostate.

10:02

I do rely it, rely on it in the breast, so I'm gonna

10:06

ignore the hue or shade of red 'cause I don't think

10:08

it's reliable And all I'm doing is looking for hyper

10:12

vascularity in the PZ to see if that corroborates my

10:17

tumor, but also helps me decide how vascular it is.

10:22

For the more vascular it is earlier on,

10:24

the more likely it is to be aggressive.

10:27

The more intense the signal intensity on diffusion

10:31

weighted imaging, the quicker the rise from low

10:34

B values to high B values, the more aggressive

10:37

the tumor, and the more hypo dense, or hypo

10:42

intense, the signal is on the ADC parametric

10:45

map, the more likely the tumor is aggressive.

10:49

Let's take a look at our ADC map, and here it is.

10:54

Not quite as black as I would have

10:55

expected, but it is present, it is there.

10:59

And again, the blacker it is, the more

11:01

aggressive it's going to be on the ADC map.

11:05

3 Tesla study, posterolateral tumor.

11:09

This would be a PI-RADS 4/5.

11:12

Depending upon whether we

11:13

measure it out at less than 1.5 or

11:16

greater than 1.5.

11:17

228 00:11:18,599 --> 00:11:20,440 The reader said slightly greater than 1.5.ss

11:20

230 00:11:21,000 --> 00:11:23,640 I measured it, I got slightly less than 1.5.

11:23

232 00:11:24,030 --> 00:11:26,289 So I'd be reading it as a PI-RADS 4.

11:26

The reader read it as a PI-RADS 5.

11:29

This looks like a moderately aggressive lesion.

11:31

Diffusion image, not that bright.

11:34

ADC map, not that dark.

11:37

Hypervascularity, some.

11:39

In fact, moderate.

11:41

Compare this to our 1.5

11:42

comp vignette.

11:44

That concludes this one.