0:01

Okay, next we're going to focus

0:02

on bright blood cine imaging.

0:04

And so here on the screen, what you see

0:07

are two sets of bright blood cine images.

0:09

The left-hand side is a short axis cine image

0:12

at the mid cavity of the left ventricle.

0:14

And the right-hand side is a four-chamber cine image.

0:17

And as I mentioned in the protocol

0:20

section, the bright blood cine imaging

0:22

is really the workhorse for cardiac MRI.

0:24

And for some indications,

0:26

really, the only sequence that we need to use, let's

0:30

say somebody is just getting an evaluation of their

0:33

function, all you need is the bright blood cine images.

0:37

They're the most important for looking at wall motion.

0:39

So how do we use them?

0:41

I find it most helpful to look at the

0:44

wall itself, and if I'm evaluating wall

0:46

motion, I want to see the wall thickening.

0:49

I want to see that the thickening is nice and uniform

0:51

all the way around the entire

0:53

circumference of the heart.

0:54

And then the same thing goes for the long axis cine

0:57

images, in this case, the four-chamber cine image.

0:59

You can see that, for instance, if you look at

1:01

the septum here, the base and mid portion of the

1:05

septum doesn't really thicken a whole lot during

1:09

systole, right there, whereas the lateral wall may be

1:13

a little bit more, with a little bit more motion.

1:15

So here you see some hypokinesis,

1:17

meaning some reduced contraction

1:19

of that wall, whereas in this case, you can see

1:22

the right ventricle, that wall is moving quite

1:24

a lot better than the left ventricular wall.

1:26

So these are the types of things that we

1:28

evaluate on the bright blood cine imaging.

1:30

We're looking for wall motion, and then, which

1:33

I'll touch on later, we also use those cine images

1:37

to evaluate function and actually

1:41

quantify ejection fraction and volumes.

1:44

So bright blood, what are some other names and

1:47

terms that we hear referring to bright blood?

1:50

Well, the most generic term for bright blood imaging

1:53

is balanced steady-state free precession imaging.

1:56

And that's the sort of generic physics-based

1:59

terminology for this particular sequence.

2:01

The trade names that you're going to

2:03

hear are TruFISP for Siemens, FIESTA

2:07

for GE, and Balanced FFE for Philips.

2:11

And what it is at its heart is an ECG

2:14

gated balanced gradient echo sequence.

2:16

You can actually perform them without ECG

2:19

gating, for instance, in abdominal applications.

2:21

You may perform them without gating,

2:23

but generally when we're doing cardiac

2:24

MRI, we're always going to have gating.

2:27

And the nice thing about the bright blood,

2:28

and what makes it so useful, is it has the

2:31

highest signal-to-noise ratio per imaging

2:33

time of all currently available sequences.

2:37

And how does it do that?

2:38

Well, it uses these very, very short TRs and TEs.

2:42

And because they're so short,

2:44

you actually have residual signal left

2:46

over at the end of each excitation.

2:48

And that signal is basically recycled into the next TR,

2:52

using balanced gradients to maximize the signal.

2:55

So, basically, that's the balanced part of

2:58

the balanced steady-state free precession.

3:00

The steady-state part is that you get this

3:02

residual signal that builds up over time

3:05

and eventually reaches a steady state.

3:07

So that's the explanation, sort of the very,

3:10

very low-level explanation of the physics behind

3:13

bright blood steady-state free precession.

3:15

The thing you really have to know about

3:16

these bright blood images is that you can't

3:18

really use them for tissue characterization

3:21

because you have both a mix of T2

3:23

and T1 weighting in these sequences.

3:25

So, you will see that fluid is bright.

3:28

In this case, in the upper right

3:29

hand corner, this is the stomach.

3:31

And you can see that the fluid

3:33

in the stomach is bright.

3:34

There's also some pericardial fluid, a little

3:37

bit of pericardial fluid right here, also bright.

3:40

And then blood is bright,

3:42

you know, obviously from the name.

3:43

But then other T2 bright things such as,

3:45

you know, edema in the myocardium might be bright.

3:47

But the thing that's confusing is T1 is also bright.

3:50

So this is actually a little fat pad

3:52

in the pericardium anteriorly here.

3:54

That's bright.

3:55

If you gave contrast, that's bright.

3:58

So you can't really use bright blood images for tissue

4:01

characterization, which is their one shortcoming

4:03

and something you really need to be aware of.

4:05

What else do you need to know

4:06

about bright blood imaging?

4:07

Well, generally we use them as a cine

4:09

image with about one slice per breath hold.

4:12

If you have a really good breath holder, like a

4:14

young patient, for instance, you can usually squeeze

4:16

out maybe two or even three slices per breath hold.

4:19

But generally, when you keep them high resolution,

4:21

you're limited in the number of slices you can get.

4:23

So it can take quite a while to

4:25

get through the entire heart.

4:26

Most times if we're imaging the heart, we're going to

4:28

have about 12 slices to get through the entire heart.

4:31

And so it'll be about, you know, 12 breath holds or so.

4:34

So it can be quite a while to get through

4:36

the entire set of images you need.

4:39

The biggest limitation that we know about

4:40

for bright blood imaging is that it's very,

4:42

very sensitive to field inhomogeneity.

4:45

Those balanced gradients, you can almost

4:46

think of it like you have these balanced

4:48

gradients that are recycling the signal and

4:50

building up this steady state of signal.

4:52

Well, if you add some sort of artifacts into that whole

4:55

situation, then those artifacts also build up over

4:58

time and also kind of increase into a steady state.

5:01

So, compared to other sequences, you get a lot

5:03

of artifacts related to field inhomogeneity,

5:06

and the things that cause field inhomogeneity are metal.

5:09

So, for instance, if a patient has a device

5:11

or has a lot of iron in the liver, or maybe

5:14

even has some clips around their GE junction

5:17

because they had gastric bypass or something

5:19

like that, you're gonna see a lot of artifacts.

5:21

Another area that we see artifacts is you can

5:24

see how close the stomach is to the heart.

5:27

Well, if you happen to just, just happen to

5:28

have a big stomach, a big bubble of air in the

5:31

stomach right underneath the heart, that can

5:32

actually cause quite a lot of artifacts too.

5:34

So, in those situations where you see a

5:36

lot of artifacts because of these bright

5:38

blood images and their sensitivity to

5:40

field inhomogeneity, what do you do?

5:42

Well, we find that the spoiled gradient echo

5:44

sequences, these are the, kind of the old standard for cine imaging.

5:48

They work a lot better when it comes to artifacts.

5:51

They have less susceptibility. So for instance, in this

5:54

case, this patient actually does have a device.

5:56

You can see this black kind of void

5:59

in the upper left-hand corner of the

6:01

image, that's where the cardiac

6:02

device is on this patient's chest wall.

6:05

And then you can see the lead in their right ventricle.

6:08

But despite all of this, you can still get decent

6:11

imaging of the myocardium and the blood pool.

6:13

Something that really is not going to be

6:15

possible with your standard SSFP images.

6:18

One of the biggest reasons we use bright

6:20

blood imaging is to do quantification

6:23

of cardiac function and cardiac volumes.

6:25

And how does that work?

6:26

Well, this is

6:27

a graphical representation of the Simpsons

6:31

method, which is the method that we use to

6:34

evaluate cardiac volume, and it's very simple.

6:36

The idea here is that you use a summation of

6:39

discs, so each slice is a disc, and when you add

6:43

up all these discs, you get an entire volume.

6:46

So what happens is, with your specialized

6:48

software, you outline the disc using some sort

6:51

of endocardial tracing method, and then the

6:56

area is calculated by combining the circle that

6:59

you measure on the image with the slice thickness.

7:03

And then from that, you can get the

7:04

area and the volume of each slice.

7:07

And then those volumes are added up

7:08

together to get the total volume for the

7:11

left ventricle and the right ventricle.

7:12

And then, basically, you repeat this process

7:14

for both end systole and end diastole.

7:17

That'll give you your end systolic and end

7:19

diastolic volumes, and then from that, you can

7:21

calculate an ejection fraction, and that's it.

7:24

It's pretty simple, actually.

7:27

All right, so let's next talk about different

7:29

limitations that we encounter with cine imaging.

7:32

So this is not specific just to SSFP imaging, but

7:35

rather limitations of just Cine Imaging in general.

7:39

And the way cine imaging works and the

7:41

simplest kind of way to look at it is that

7:44

you're acquiring multiple bits of information

7:47

from separate heartbeats and combining them

7:50

together to make a whole, an entire image.

7:52

So for instance, if you get 30 frames in

7:55

your cine image, then you get a little

7:57

bit of frame 1 through, you know, 30.

8:00

The very, you know, first, say, set of pixels on one

8:03

heartbeat, and then a few more on the next heartbeat,

8:06

and a few more on the next, and then when you

8:07

combine it all at the end, you get a complete image.

8:10

So, the limitations of cine, basically, are due to

8:13

the fact that you're relying on this steady heart

8:16

rate and consistency from heart rate to heart,

8:18

from heartbeat to heartbeat to get good imaging.

8:21

So if you're combining all this data from multiple

8:24

heartbeats and your heartbeats are changing from

8:26

one beat to the next, then you can get blurring.

8:29

So if you take a look at this figure here in the

8:31

upper right-hand corner of the slide, this is

8:34

an example of what can happen with arrhythmia.

8:36

So if you look at the PQRS complex

8:38

here, there's a certain distance

8:40

between this P wave and the next P wave.

8:43

And then when the following P wave comes up,

8:45

it's actually coming up a little bit early.

8:47

And so this part of the image data, which is in

8:50

late diastole for this particular heartbeat,

8:53

is actually now in systole for the next heartbeat.

8:56

So you end up combining image data from both

8:58

diastole and systole together, and you get blurring.

9:02

A similar situation can happen when a patient

9:04

is having difficulty with breath holding.

9:07

In this case, you can have a regular,

9:09

nice, regularl spaced heartbeats with

9:11

a consistent P wave after your QRS.

9:14

But if the patient is inhaling and then exhaling,

9:18

you're going to have different parts of the heart

9:21

in the image from one acquisition to the next.

9:25

So, for instance, on heartbeat A, maybe

9:27

you're looking at the top of the heart.

9:29

But then, if the patient takes in a deep

9:31

breath on heartbeat B, you're going to be

9:34

looking actually at the ascending aorta.

9:36

And then, you know, vice versa, and so up and

9:38

down, up and down as the patient breathes.

9:40

Again, if you have pieces of information

9:43

from different heartbeats that are being

9:45

combined and they aren't exactly the same,

9:48

then you're going to get some blurring.

9:50

So how do we counteract this in cine imaging?

9:53

Well, the most commonly used

9:55

approach is to use real-time imaging.

9:57

It can be used in difficult cases.

9:59

The problem with real-time imaging is that you

10:01

do take a hit on spatial and temporal resolution.

10:05

So you can think of real-time imaging

10:06

basically as like a video camera.

10:09

So rather than combining different pieces of data from

10:11

multiple heartbeats, you're actually just turning on

10:14

the camera and watching what happens in the heart.

10:17

In order to make that happen, you need to acquire

10:19

the images really, really, really quickly.

10:21

And so that's where the spatial and temporal

10:23

resolution sacrifices have to occur.

10:25

You need to basically reduce the spatial

10:27

resolution, reduce the temporal resolution,

10:29

so that you can cram in all the information you can

10:32

in order to get the images as fast as possible.

10:35

And so, I'm going to show you

10:36

an example on this next slide.

10:38

So on the left-hand image here, this is cine

10:40

imaging of somebody with poor breath holding,

10:42

and you can see all the blurring that's occurring,

10:45

especially at the end of systole there.

10:48

You can see diastole looks okay, but then as the

10:51

contractions happen, there's a lot of blurring.

10:53

It would be really hard if you were doing

10:55

contouring and trying to define the wall in

10:59

order to calculate ejection fraction or volumes.

11:02

It would be very difficult in this case.

11:04

So here's the solution.

11:06

Real-time imaging.

11:08

You can see that there's not

11:10

quite the level of detail, right?

11:12

It just looks a little blurry.

11:14

The edges aren't quite as sharp.

11:17

But the nice thing is you don't have that same

11:19

blurring that you have with the other image.

11:22

So you don't have the smearing effect.

11:24

The actual resolution is lower, but you do see nice

11:29

sharp endocardial borders here, and you could, in fact,

11:32

use that software to calculate an ejection fraction.

11:37

You can see this patient actually was in bigeminy.

11:40

If you remember back from med school, that's

11:42

when people have two abnormal heartbeats in a row

11:45

that just keeps repeating over and over again.

11:47

So you can see there's a one, there's a two systolic

11:51

contractions, which can be a problem for the

11:54

standard cine software to figure out, and that's

11:57

why we have the blurring on the breath hold imaging.