0:02

Hello and welcome to Noon Conference hosted by MRI Online. In response to

0:08

the changes happening around the world right now and the shutting down of

0:11

in person events, we have decided to provide free Noon Conferences to all

0:16

radiologists worldwide. Today, we are joined by Dr. Juan Carlos Batlle.

0:21

Dr. Batlle is the Chief of Thoracic Imaging for Baptist Health South Florida

0:27

and Miami Cardiac and Vascular Institute, as well as Associate Professor

0:33

at FIU College of Medicine. He serves as Cardiac MR Course Director for

0:38

the American College of Radiology Education Center. A reminder that there

0:42

will be a Q&A session at the end of the lecture,

0:45

so please use the Q&A feature to ask your questions and we will

0:49

get to as many as we can before our time is up.

0:52

That being said, thank you all for joining us today. Dr.

0:55

Batlle, I'll let you take it from here. Hopefully, everyone can see my

0:59

screen. Thank you for the kind introduction and hopefully, this is helpful

1:03

for you. I'll kind of get into the meat of it because we

1:05

do have a lot of material and I'd like to be able to

1:07

answer your questions. So as mentioned, I do co teach. I'm the Director

1:11

for the Cardiac MR Course at the ACR.

1:14

This year, most of it has been canceled, but we are anticipating October

1:17

15 through 17. So this is kind of

1:19

a taste for what would be offered in person in Reston, Virginia,

1:23

and Northern Virginia at that course. The first cardiac MRI that was published

1:28

was about 40 years ago now. And you can see from the images

1:32

below that it was relatively low resolution, but it was ECG gated. So

1:36

we did still the motion of the heart in the so called black blood

1:40

technique, showing that you could see chamber sizes and anatomical abnormalities.

1:46

But we really didn't get a precise look at the heart and contrast

1:48

that with the 21 Tesla MRI on the upper right of a rat

1:53

heart. This is here in Florida at a research lab and a research

1:56

magnet, not a human sized magnet, showing that we can really exquisitely

2:01

demonstrate the myocardial architecture at the fiber level with MRI and

2:06

do tractography. So we've come a really long way. And so I'm going to try

2:10

to pitch in the middle for something where we are clinically relevant,

2:13

things that we're doing today, and to give you a flavor of the

2:16

kind of entities that we commonly interact with in cardiac MRI.

2:21

First thing, we use a standard MRI machine, that's our Philips Achieva 1.5T.

2:26

You don't need a special magnet. Really the coils that you'd like are

2:31

kind of the 16 32 channel coils for better coverage of the heart

2:37

and better sensitivity. We do ECG gate, and sometimes we do add respiratory

2:42

gating so you can add a bellows to the abdomen so that you

2:45

can see if the chest is excursing or not. This is an example

2:50

of our 32 and 5 channel coils. We use 32 whenever possible,

2:54

5 channel is for smaller patients, pediatric, or patients where the 32 channel

3:00

wouldn't fit for whatever reason due to the geometry of the patient interacting

3:04

with the magnet. We're looking to gate to the ECG cycle,

3:08

and one of the challenges there is, you have a very small amount

3:10

of time to image the heart and still the beating of the heart.

3:14

So if you imagine an RR interval being one heartbeat,

3:18

you want to kind of... Those little blue bars kind of give you

3:20

the idea of when we're collecting data. Whether it's CT or MRI,

3:24

you've got the shutter open like a camera for a certain amount of

3:27

time, and the longer you leave that shutter open, the more blur there

3:31

is because, of course, as you go from one R wave to the

3:33

next R wave, the heart is moving and then returning to its initial

3:37

state. So the more your shutter is open, the more of that movement

3:41

you capture, and therefore you're not getting a focal end diastolic image.

3:45

You're getting a mid to late diastolic image or an early to late

3:50

diastolic image if you leave it open. So that's our challenge in cardiac

3:54

imaging. And one of the things that we do to address that is,

3:57

segmented acquisition of K space. We generally are not looking to acquire

4:01

all of the K space for one slice in one heartbeat. What we

4:05

do is we get bits of the cardiac cycle at a time.

4:09

And so we construct, in this image, you can see the upper half

4:13

of the image, the interior part of the image.

4:16

We get that in earlier portion of the cardiac phase and systole.

4:20

And then instead of trying to get that same slice of the heart,

4:26

the entire image to get the posterior, instead, we just wait a little

4:30

bit and get that same anterior slice. Again, we segment K space to

4:34

only get enough lines of K space so that we aren't widening that

4:37

shutter opening. We're not taking too much time to acquire that image and

4:42

creating blur. And then after multiple heartbeats, you can fill in the posterior

4:46

parts. Here on the second heartbeat, we're getting the second eight lines

4:50

of K space that fill in that posterior part of the image in

4:53

the systolic frame. And then we wait a little bit and get the

4:55

posterior slice of the later systolic or the diastolic frame.

5:00

That can be a challenge with high heart rates where you have less

5:04

time to image and the heart moves faster in the same amount of

5:07

time. But we do have other approaches to that, I'm not going to

5:12

get too far into the physics of our protocols.

5:16

Cardiac planes are different from the ones that you're used to.

5:18

So we go from axial scouts to more cardiac type planes.

5:23

So we'll drop a plane parallel to the septum going down the barrel

5:26

of the left ventral to get what we call kind of a pseudo

5:29

two chamber view. The two chambers being the left atrium and the left

5:32

ventricle. It's not quite the two chamber view that we'd like because we

5:35

have to tilt it in the other plane. And so how do we

5:37

accomplish that? Well, we take that pseudo two chamber view and we get

5:42

a line that's perpendicular to the septum this time.

5:45

And that gets us, again, a pseudo short axis view. And these beating

5:48

heart images are bright blood, steady state free precession images is the

5:53

generic term. You may have heard terms like FIESTA or true FISP. Each

5:57

vendor has its own different way to name these, so it can be

6:02

challenging. And so a bright blood cinematic or SSFP image is kind of

6:06

the generic term for that. We're able to see the heart contract and

6:09

the walls move, the size of the chamber and so forth.

6:13

From that pseudo short axis view, we can get to a true four

6:16

chamber view. So we're starting to add some cardiac images, cardiac planes.

6:22

And here we have a nice four chamber view from which we can

6:25

then generate a true short axis. So we get images that are,

6:29

again, perpendicular to the septum. We go all the way up and down

6:32

and then we get a short axis view, ultimately getting something like this.

6:36

In this case, nine images showing the beating of the heart from about

6:40

the valve plane, the atrioventricular valve plane to the apex. And we can

6:44

evaluate the heart at any of those levels. So that's the goal of

6:47

your cinematic images. And then we'll get some tissue characterization images

6:50

as well. One special note is to talk about dephasing. So the bright blood

6:55

images, when the blood doesn't look too bright in the bottom light,

6:59

in the bottom...