0:02

Hello and welcome to Noon conferences

0:03

hosted by MRI online.

0:05

In response to changes happening around the world right now

0:08

and the shutting down of in-person events, we've decided

0:10

to provide free daily noon conferences

0:12

to all radiologists worldwide.

0:14

Today we're joined by Dr.

0:16

Ron Hu for a lecture on neuro imaging of epilepsy. Dr.

0:19

Hu is assistant professor of Radiology

0:21

and Imaging Sciences at Emory University,

0:24

associate Program Director of the Neuroradiology Fellowship

0:27

and Director of Stroke Imaging.

0:29

His clinical interest include neurovascular imaging epilepsy

0:32

and functional MRIA reminder, there'll be a q

0:35

and a session at the end of the lecture, so please use the q

0:38

and a feature to ask your questions,

0:39

and we'll get to as many as we can before our Time's Up.

0:42

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

0:44

I'll let Dr. Hu take it from here.

0:47

Hello everyone again.

0:48

Um, thank you for joining this new conference.

0:51

Uh, my name is Ron Hu.

0:53

I'm from University, uh, Emory University in Atlanta,

0:55

and today I'll be speaking about neuroimaging in epilepsy.

1:01

Um, have I'm involved in a clinical trial on laser ablation

1:05

for temporal lobe epilepsy,

1:06

but have no financial involvement.

1:09

Uh, so the objectives today is to, um,

1:13

talk about why do we image epilepsy, so the rationale

1:15

for imaging, uh, when to image in epilepsy,

1:19

and then go over some of the common, uh,

1:21

pathologies that cause epilepsy.

1:24

Uh, mostly focusing on adult pathologists, uh,

1:27

including temporal lobe abnormalities, malformations

1:29

of cortical development, uh, epileptogenic tumors, uh,

1:33

just a couple of syndromes.

1:34

And then, uh, a, um, kind of a growing kind of a

1:40

pathology, uh, in which is previously underrecognized.

1:44

And then finally, we'll, uh, have some time for questions

1:47

and answers, and please type those into

1:49

the, uh, q and A box.

1:52

So, uh, why do we image in epilepsy?

1:54

Uh, first, let's go over the terminology here.

1:57

So, a seizure is any kind of disturbance,

2:00

electrical activity in the brain.

2:01

And then when you have two unprovoked

2:04

seizures within 24 hours,

2:06

or have one in the, you know, a high likelihood

2:09

of another one that qualifies as a epilepsy syndrome.

2:12

So, epilepsy affects about three

2:15

and a 3.4 million people in the us

2:18

and about the majority of them actually have unknown cause.

2:21

So when patients present initially with seizure, you know,

2:24

we image in order to diagnose an acute cause of seizure,

2:27

such as infection or trauma, intracranial hemorrhage,

2:31

and also to exclude, you know, other, um,

2:33

pathologies like brain tumors.

2:35

Um, so in terms of epilepsy though, uh, we image really

2:40

to kind of guide treatment.

2:41

So, uh, we know that up to a third of patients

2:44

with epilepsy are refractory to medication.

2:47

And nowadays there's, uh, very effective epilepsy surgery

2:51

and even neuromodulatory therapies that can be targeted

2:54

to people with refractory, uh, epilepsy.

2:57

So, and, uh, papers have shown

2:59

that surgical success is actually associated

3:01

with the recession of the focal, uh, seizure zone.

3:05

So being able to identify structural cause

3:07

for focal epilepsy is,

3:09

is very important for treatment planning.

3:13

This is, um, recent abstract,

3:15

I guess it's impressed right now in the H-J-A-C-R, uh, done

3:19

by one of our residents who looked at utilization of imaging

3:22

for seizure and epilepsy in the er.

3:24

We saw that, uh, testing in terms of both EEG, CT

3:28

and MRI for seizure

3:30

and epilepsy has increased considerably between 2006

3:33

and 2014 in the us, uh, going up, uh, you know, 400%

3:38

for MRI in adult patients.

3:41

And this has outpaced the change in number of ED visits.

3:44

And looking at the aggregate data, we found that about nine

3:47

to 12% of, um, these, uh, emergent, uh,

3:51

CT imaging actually found some acute pathology, um,

3:54

in the ed, but in patients with known epilepsy disorder, um,

3:59

43% of them actually had imaging in the er,

4:02

but only 3%

4:03

of those cases actually showed anything that was actionable.

4:06

So, patients with first, you know, uh, new onset CT

4:10

or new onset seizures presented to the ed, you know,

4:13

the ELs about 10% for ct.

4:16

Uh, but then if you have a known epilepsy disorder, uh,

4:19

the yield for imaging in the emergent setting is much less.

4:23

So, um, today I'm gonna focus mostly on CT and MRI.

4:27

Uh, I know, you know, these imaging methods are part of a,

4:31

you know, um, comprehensive workup.

4:33

So that would include pet, maybe even MEEG

4:36

and the invasive monitoring.

4:37

Uh, but for today's lecture,

4:39

we will focus on cross-sectional imaging, CT and MRI.

4:42

Uh, CT is fast

4:43

and available in the first line, uh, imaging tool.

4:47

In the acute setting. It can be used

4:49

for detecting calcifications, uh, such

4:51

as in infectious diseases like neurotic psychosis, uh,

4:55

even cavernous malformations.

4:56

Um, and then osseous defects, uh, for, from trauma

5:00

or, uh, cephaloceles.

5:03

Uh, sometimes CT profusion is done, uh,

5:06

because when they, uh, present,

5:07

we don't know if it's in seizure or stroke.

5:10

So, uh, in the, um, acute setting,

5:13

sometimes CT profusion is done

5:14

and would show hyper profusion in the ictal

5:17

or immediate ictal, uh, period.

5:20

And then finally, uh, we do CT for surgical planning

5:22

and localization, uh, to plan for electro placement, as well

5:26

as, uh, you know, ablations or surgery.

5:29

In terms of, uh, MRI, uh, it definitely is higher costs

5:33

and run into issues with compatibility.

5:35

And a lot of people with, uh,

5:37

epilepsy actually have implanted devices like, uh,

5:40

vagal nerve stimulators and, uh, neuromodulatory devices.

5:44

So compatibility becomes an issue.

5:47

Uh, we typically image patients with epilepsy

5:49

with a dedicated epilepsy protocol,

5:51

preferably at higher strength field strength, like three

5:54

or even nowadays at 70.

5:57

Uh, normally, uh, if the patient is, uh, kind of a older

6:01

or middle age adult, we would give contrast, well,

6:04

brain tumors that are infectious inflammatory processes.

6:08

Um, we sometimes do MR spectroscopy to differentiate

6:10

between the tumor or cortical dysplasia, uh,

6:13

and we sometimes do MR profusion for that as well.

6:16

And then, uh, we, uh, often perform both fm I both task base

6:20

and resting state FMRI for presurgical planning.

6:24

So our protocol here, uh, is, is outlined, uh, to the left.

6:29

Uh, but you know, you can substitute a lot

6:31

of these protocols for, you know, 2D for 3D um, variants.

6:35

So the 2D uh, sequences typically have higher contrast

6:39

and noise ratio, higher in plain resolution,

6:41

but they're thicker slices.

6:43

The 3D um, variants typically have kind of isotropic, uh,

6:48

one millimeter or less resolution

6:49

that you can reformat in any plane,

6:51

but they typically suffer from, uh, a little bit, uh,

6:55

in terms of, uh, signal to noise.

6:58

And when we do a image of hippocampus, it's important

7:01

to do o bleak coronal as opposed to a straight coronal, uh,

7:05

which means that, uh, we orient the, uh, plane

7:07

to be perpendicular to the axis of the hippocampus in order

7:11

to view it in profile.

7:15

In terms of the different causes of epilepsy, in terms

7:18

of structural causes, uh,

7:20

the most common would be mesial temporal sclerosis in adults

7:23

is a focal cause of epilepsy.

7:25

And we also have various tumors.

7:27

Um, this is not an exhaustive list,

7:29

but now any tumor can cause, uh, seizures.

7:33

But these are kind of typically low grade tumors we think of

7:36

as being epilepsy causing tumors, including ganga gliomas,

7:39

DNA and PXA, uh, developmental abnormalities in terms

7:43

of malformations of critical development.

7:45

Uh, various neuro cutaneous, uh, syndromes, uh,

7:49

vascular malformations like cavernous malformations, even

7:52

AVMs sometimes can cause seizures.

7:55

Um, and increasingly recognized, uh, is, uh,

7:58

autoimmune encephalopathy, which I'll show some cases later.

8:01

And of course, any type of infection,

8:03

trauma, and infarction.

8:05

Uh, typically infarction doesn't cause, um,

8:07

seizures in adults, typically.

8:09

That's a presentation in kids.

8:13

So now I'm going to, for the next few minutes,

8:15

focus on temporal lobe epilepsy.

8:17

Uh, first focusing on the imaging appearance

8:20

and, uh, classification of mesial temporal sclerosis.

8:23

Uh, then some of the secondary, uh, kind of, uh, uh,

8:27

imaging features, including amygdala enlargement,

8:30

hippocampal mal rotation,

8:32

and, uh, abnormal temporal lobe signal.

8:34

And then finally, show a couple cases of, uh, autoimmune

8:37

and, uh, infectious encephalitis.

8:41

So, um, kind of going back to neuroanatomy here.

8:44

So what, uh, makes up the medial temporal lobe

8:47

or medial temporal lobe,

8:48

depending on your preference, it's the same thing.

8:51

So basically as three main parts, uh, the amygdala,

8:55

hippocampus, and per hippocampal jars.

8:57

Let's see if this plays here.

9:01

So, um, the amygdala's, you know, anterior portion

9:04

of the medial temporal lobe shaped like an almond.

9:07

Um, and then, uh, you have the onca, which,

9:11

which is the most medial portion of the amygdala.

9:14

Uh, then going back, we start to see the interdis

9:18

of the, uh, hippocampus.

9:19

So this is a hippocampal head,

9:21

which has this characteristic undulating appearance.

9:24

And then as you move back further, uh, you'll, uh,

9:28

see the hippocampal body, uh, which technically kind

9:31

of on the chrono plane begins at the level

9:32

of the red nucleus.

9:34

You can kind of see a faint outline of that there.

9:36

And as you move back further back at the level of the, uh,

9:40

superior inferior ulus, you have the beginning

9:43

of the hippocampal tail.

9:44

And then as you go back further, um, you have the FIA here,

9:48

which you eventually, uh, you know, becomes the fornix

9:52

and kind of loops back forward again.

9:56

Uh, so kind of inferior to the hippocampus.

9:59

We have the, uh, para hippocampal gys, which is here,

10:03

and that's separated from the fusel form gys

10:06

by the collateral sulcus.

10:08

Uh, so this is the collateral sulcus.

10:10

Here we have the gys,

10:13

and then more laterally, we have the superior, middle,

10:16

and inferior, uh, temporal gyri.

10:21

Okay. So, uh, mesial temporal sclerosis, uh, as I mentioned

10:25

before, is the most common cause

10:27

of focal epilepsy in adults.

10:29

Uh, the imaging findings are threefold,

10:31

so you have high signal, which reflects gliosis

10:34

of volume loss, which reflects atrophy

10:37

and loss of internal architecture.

10:39

So the internal architecture, uh, refers to what I mentioned

10:42

before, the, you know, in interdis of the hippocampal head,

10:47

but also this kind of, uh, strip you visualize on most MRIs,

10:50

especially three T or above.

10:52

You'll see this kind of thin linear, uh,

10:55

dark signal on T two, which we is a white matter tract.

10:58

So that's the stratum radi atom, that's a part

11:01

of the white matter tract that goes into the hippocampus.

11:03

So you should be able to recognize that, uh,

11:06

on any good quality MRI, uh,

11:08

better in three T than in 1.5 T.

11:11

However, even with, you know, current imaging techniques,

11:14

you know, 20 to 30% of, uh,

11:16

MTS can actually be m MRI negative.

11:19

So, um, there are papers that show that, you know,

11:22

in these MRI negative cases 70 can be helpful.

11:26

And, and we know that from a surgical literature,

11:29

that the postoperative outcome correlates with the kind

11:32

of degree of abnormality and MRI.

11:34

So if you have a definite MTS, you know,

11:36

the surgical outcome for treating

11:38

that is, is gonna be better.

11:41

There is a classification system for MTS, which, you know,

11:44

we don't really get into, uh, as radiologists, it's more

11:47

of a pathologic, uh, subtype, uh, system.

11:50

The type one is the most common,

11:51

which is neuronal cell loss,

11:53

and gliosis predominantly involving the ca one

11:56

and ca four regions.

11:57

Um, so here and here and type two and three are less common.

12:02

Uh, type two is a ca one predominant neuro node loss,

12:05

so more kind of the lateral portion of the hippocampus.

12:08

And then C uh, type three is a ca four,

12:11

and D jar is predominant volume loss.

12:16

Uh, so this is an example of what I showed before.

12:19

So, uh, typically, um, I like to use the T two

12:22

to look at the kind of volume, um,

12:25

and the internal ex architecture.

12:27

So here on the, uh, left side here, lemme just change

12:31

to the laser pointer.

12:33

So Le left side of the patient, we have, uh, atrophy

12:38

of this hippocampus with loss of an internal architecture.

12:41

And then the flare shows that it's hyperintensity signal.

12:47

So there are various secondary findings of, uh,

12:51

mesial temporal sclerosis, uh,

12:53

enlargement of the temporal horn.

12:55

Uh, you can have atrophy of the perim capal jars, widening

12:59

of the collateral sulcus atrophy of the,

13:02

its later mammary body here, uh, as well as, you know,

13:05

the whole, uh, pape circuit.

13:06

So the fornix, uh, even involving the thalamus.

13:10

Uh, you can also have either increasing

13:13

or atrophy of the amygdala, which is a, uh, kind

13:15

of an under-recognized sign, no show later.

13:20

Uh, often, uh, we have dual pathology, which means

13:24

that there is, uh, mesiotemporal sclerosis,

13:26

but it's actually secondary to a different, uh, kind

13:30

of a seizure, uh, provoking, uh, pathology.

13:34

So it occurs about 15% of cases.

13:37

This is a case of the left MTS that actually had a, uh, uh,

13:42

focal cortical dysplasia.

13:47

This is a, uh, case of, um, MTS on the left,

13:50

but there's an underlying, uh, kind

13:52

of a developmental venous

13:54

or, uh, underlying, uh, cavernous malformation.

13:57

Here you see the complete hemo citrin rim of the cab mount,

14:00

and you see associated DVA, um, beating, uh,

14:04

away from that cab mount.

14:05

So it's important when you look at MTS

14:07

and not get kind of, uh, satisfaction of surge

14:10

and stuff right there, it's important to kind of look

14:12

through the entire temporal lobe, even outside

14:14

of the temporal lobe, to look for dual pathology,

14:17

because it really affects treatment.

14:21

Um, sometimes you can get bilateral MTS,

14:24

so it's observed in about three to 10% of cases.

14:27

Uh, so here's an example

14:28

where you have atrophy in hyperintense signal

14:32

of both hippocampi.

14:34

Uh, some usually though the symptoms,

14:36

you know, uh, is unilateral.

14:38

So, um, you know, one of these is, uh, is, uh,

14:43

causing the seizures, but it's very hard to tell by imaging.

14:45

So, uh, they'll probably need more invasive, uh, testing.

14:51

Nowadays, I think a lot of institutions have started using

14:54

kind of volumetric analysis, uh,

14:57

to assist in quantifying the, uh, hippocampal size.

15:00

And here we use, uh, a neuro con,

15:02

but there are others available.

15:04

And there was a paper actually looking at, you know,

15:06

comparing the sensitivity

15:08

and specificity of, uh, neuroco versus neuroradiologist.

15:11

And it was, it was actually neuro con one,

15:13

but it was statistically not significant.

15:15

Um, but this is only looking at volume loss.

15:18

So, uh, it's important for us as radiologists to kind

15:21

of check the neuro con for any type

15:23

of quantitative analysis, make sure the segmentation is

15:25

correct, and to also, um, correlate

15:28

with other imaging findings in clinical history.

15:32

And, uh, the normal, uh, usually these kind

15:34

of software spits out a number, which is asymmetric index.

15:37

It's basically, um, uh, the percentage kind

15:42

of asymmetry between the two sides.

15:44

And, uh, in normal, uh, subjects, you can expect an index

15:49

of 2.1%.

15:51

So there is some natural variation in terms of size

15:53

of the hippocampus, and it's important not

15:55

to call these too tightly.

16:00

I mentioned, uh, 70 T before.

16:03

Um, there, this is a paper I actually know pretty old now,

16:06

nine years ago, that, um, looked at 70, uh, in MTS.

16:11

And, uh, you know, 17 is becoming more popular.

16:14

There's clinically approved, uh,

16:16

devices now from the major vendors.

16:18

So we'll beginning to see this more often.

16:21

Uh, in terms of the, uh, what you can see,

16:23

you can see the interdis very clearly.

16:25

You see the striation much more clearly on seven

16:28

T than on three T.

16:30

Um, in this group of, um, a small group,

16:32

they actually observed hippocampal mal rotation in

16:35

four out of 11 patients.

16:37

And I'll, I'll mention what that is later.

16:40

Uh, in, in temporal lobe epilepsy, they found that there was

16:43

posity of the interdis of the, uh, hippocampal head in all

16:47

of their patients with, uh, temporal lobe epilepsy.

16:50

Uh, even they saw that in three out

16:52

of eight on the contralateral side,

16:55

and they saw, they saw a loss of that striation

17:00

in terms of hippocampal marrow rotation, basically

17:02

what it is, is a, uh, it's, uh, kind

17:05

of a abnormal morphology of the, uh, hippocampus

17:09

where normally it's kind of lying on its side,

17:12

but in the abnormal, uh, side, it's kind of more upright.

17:16

And as a kind of a, a secondary feature,

17:19

the collateral sulcus, which is normally more kind

17:22

of at a 45 degree angle, it's more, uh, upright as well.

17:27

You have abnormal vertical and deep collateral sulcus.

17:30

Um, there's some debate, uh, in, in terms of the cause

17:34

of relationship between, uh, this finding and seizure.

17:37

Uh, I think most people now believe the

17:39

CBA developmental variant.

17:41

Uh, it's not directly related to mts,

17:45

but can be a sign of kind of a general developmental, um,

17:50

abnormality, you know, uh,

17:51

that signals something went wrong in the, uh, formation

17:54

of the temporal lobe, but it's not directly related to MTS.

18:01

Uh, Amy, the amygdala enlargement is something

18:04

that's also kind of recently recognized, um,

18:07

as a part of MTS.

18:08

So, um, you can have a am amygdala enlargement in

18:12

oscillation, uh, with increased signal, um,

18:15

and on pathology, you can have, basically, it shows, uh,

18:19

a type of dysplasia with hypertrophic neurons

18:22

and atrophy, uh, gliosis, uh, often though, uh,

18:25

in intraoperative recordings, even though the, uh,

18:29

findings in amygdala, they also record

18:31

that the seizure is actually coming from the hippocampus.

18:34

So, um, it's kind of a

18:37

unknown whether this is a secondary finding of, uh, MTS

18:41

or a primary cause of, uh, you know, distinct type

18:44

of mesial Temporal sclerosis,

18:48

reportedly it's less associated with febrile seizures

18:50

and has a later onset than the classic MTS

18:53

that's only involved in the hippocampus.

18:56

Uh, important to recognize though there, uh, kind of mimics

18:59

of, uh, this, uh, amygdala enlargement, uh, in the immediate

19:03

kind of postictal period, you can have enlargement of the,

19:06

uh, amygdala in the hippocampus.

19:09

Uh, you can also have low gliomas that occur in this region,

19:13

uh, and, uh, even encephalitis.

19:15

So, um, kind of important to recognize those as well.

19:21

Anterior temporal lobe signal is, uh, is another kind

19:24

of subtle secondary feature that you may observe, um,

19:28

is reportedly observed, uh,

19:30

in the ipsilateral anterior temporal white matter in the

19:33

third of patients with, uh, me,

19:35

medically refractory temporal lobe epilepsy.

19:37

And what it is basically is kind of this, uh,

19:40

hypertense signal in the white matter in the, uh,

19:43

temporal pole, or adjacent to the hippocampus in the,

19:46

in the collateral white matter.

19:48

And you can visualize better, uh,

19:50

if your institution does a, uh, kind of a,

19:52

a double inversion recovery.

19:53

So we do a double inversion recovery sequence here,

19:56

which kind of has two saturation pulses, saturates the, uh,

20:00

CSF, as well as, uh, the, uh, my kind

20:03

of the myelin in order to accentuate the, uh,

20:06

abnormalities in a white matter.

20:08

So DIR tend to be more accurate, uh,

20:10

to lateralize this finding.

20:14

This is, uh, an example of, uh, autoimmune encephalitis.

20:17

Basically, you have kind

20:19

of the enlargement in abnormal signal of the, uh,

20:22

both the left and right, uh, ma, the left and hippocampus.

20:26

Um, these, uh, you know,

20:28

tend can have some left menal enhancement if the image these

20:31

over time, they tend to develop atrophy

20:34

and scoliosis over time.

20:36

And as you know, you know, there are various causes

20:37

of autoimmune encephalitis, uh,

20:40

vari various antibodies like the anti NMDA receptor

20:44

antibody, as well as multiple tumor associated

20:47

autoimmune causes.

20:51

This is a case, um, of a infectious, um, uh,

20:56

temporal lobe seizure.

20:58

Uh, it's actually presented to us,

21:00

you know, as a stroke alert.

21:02

So the patient, uh, went through this whole C-T-C-T-A

21:06

and CTP, and on the CTA we saw there was some hypo density

21:10

in the right insula, uh, even sub insular white matter.

21:15

And the, uh, preliminary report was

21:17

that this may be a right MCA stroke,

21:19

but if you kind of, uh, look more closely,

21:22

there's actually hypo density also involved the right

21:25

amygdala, which is in a different vascular territory.

21:28

So the amygdala in the very tip

21:30

of the hippocampal head is actually supplied

21:33

by the anterior choroidal artery,

21:34

which arises from the ICA is not from the MCA.

21:38

And then in vascular imaging,

21:39

we saw there was no large vessel occlusion.

21:41

So the ICA was not occluded.

21:43

The MCA was widely open,

21:45

and there's actually kind of a increased, uh, prominence

21:49

of the kind of particular branches on the right.

21:51

And on profusion there was increased blood volume

21:53

and increased, uh, blood flow.

21:56

So this was a case of herpes encephalitis, uh,

21:59

which has shown better on MRI.

22:01

You see, the extent

22:02

of the abnormal signals is more kind of broad.

22:05

It involves the frontal lobe, temporal pole, and insula.

22:10

There was associated left meningeal enhancement.

22:12

There was cortical restrict diffusion

22:14

that reflects probably recent seizure activity.

22:17

And if you pay attention even more closely,

22:20

there is layering, uh, you know,

22:23

material within the occipital horn

22:25

that could be biogenic material or some blood product.

22:28

So, um, you know, uh, we should always be aware

22:31

of these kind of, uh, pathologies,

22:34

even when they present us, is not seizures

22:36

'cause uh, the treatment is, uh, is very different.

22:41

So, moving on to, uh, malformations of cortical development,

22:45

um, this is a pretty broad, uh, category of, uh,

22:49

diseases which can, um, you know, uh, result from the rest

22:53

of any, uh, step of the normal, uh, kind of, uh, migration

22:58

and, uh, organization of, uh, uh, neurons

23:02

as they migrate away from the germinal

23:04

matrix toward the cortex.

23:06

So when you have abnormal proliferation

23:08

or apoptosis, you can develop microcephaly, heme, cephalic,

23:12

or type two cortical dysplasia.

23:14

Uh, if you have a rest of dys normal, uh, migration,

23:18

you can get heterotopic, uh, gray matter.

23:21

Uh, so basically you can get per ventricular, uh,

23:25

or subdermal heterotopia.

23:27

You can get, uh, subcortical heterotopia or lissencephaly

23:31

or smooth brain.

23:33

And then finally, when you have abnormal cortical

23:35

organization, uh, you have, you can develop poly micro jia,

23:39

uh, sli, and for some reason, type one

23:42

and three FCD, uh, are funded under a category

23:46

of abnormal cortical organization.

23:48

Uh, but type two is abnormal apoptosis.

23:54

So, uh, focal cord dysplasia,

23:56

there is actually three types as well.

23:58

Uh, the type that we usually visualize is type two, uh,

24:02

specifically type two B, so dysmorphic

24:04

neurons with balloon cells.

24:06

But, um, you know,

24:07

type one you can see some imaging findings.

24:09

So type one refers to, uh,

24:12

just abnormal cortical lamination.

24:14

Uh, on imaging, you'll see some subtle blurring

24:17

of gray white junction, uh, but less so than type two.

24:20

You can have kind of segmental

24:22

or lower atrophy with loss

24:24

of the regional white matter volume.

24:26

Uh, you can moderate increase in T two signal

24:28

in the underlying white matter.

24:30

Uh, so these can be subtle and can be missed on imaging.

24:33

Uh, type two, uh, is the one,

24:35

typically type two B is the most common one that we see, uh,

24:38

is more common to be, uh,

24:40

more commonly found in frontal than temporal lobe.

24:45

And what you see is a cortical thickening, uh,

24:47

with increased T two signal.

24:49

There's marked blurring of the gray white junction,

24:51

and sometimes you see this, uh, trans mantal sign

24:54

of abnormal kind of white matter, um, underlying the,

24:57

uh, cortical dysplasia.

24:58

And type three is basically a, uh, SCD that's associated

25:02

with other pathologies.

25:07

So, um, this is an example of a, uh,

25:09

type two focal cortical dysplasia.

25:11

Um, this is a kind

25:13

of an abnormal gray white differentiation,

25:15

cortical thickening in the right temporal lobe.

25:17

That's important to kind

25:18

of look at these in different planes

25:20

because you don't wanna be confused by volume averaging, um,

25:24

when you look at any single plane.

25:25

So here, this one proves to be real on both planes.

25:29

And then on a T two, you can really appreciate the kind

25:32

of the cortical thickening

25:33

and even kind of abnormal signal in

25:35

the adjacent white matter.

25:38

And pathologically, this is very similar to

25:40

what you see on the histology.

25:42

Uh, you see cortical thickening

25:44

and blurring at the gray white junction.

25:48

This is, uh, an example of a transman sign.

25:51

So focal cortical dysplasia here, paramedian, uh,

25:54

it looks like frontal lobe.

25:56

Um, you have kind of increased signal in, in a blurring

26:00

of the gray white differentiation here.

26:02

Uh, but then you also see this kind of band

26:04

of abnormal white matter extending

26:06

toward the ventricle margin.

26:08

And this is, uh, the trans mantle sign.

26:10

And you can, uh, see this also better on the double inver

26:13

copy sequence at 70 T.

26:17

Uh, you can see a lot better than at three T as expected.

26:22

Uh, this was a paper, uh, that looked at 21 patients with,

26:25

uh, intractable epilepsy with negative 1.5

26:29

or three T studies,

26:30

and they found that six of these 21 patients were found

26:34

to have FCT on MRI confirmed by pathology at 70.

26:38

Um, so on the Mr I 70,

26:41

you basically can see much

26:42

better at the gray white differentiation.

26:43

Here you see this kind of abnormal, uh,

26:46

lamination gray white loss here,

26:49

which you cannot appreciate at three T.

26:54

Um, just to complete the picture.

26:56

So this is what Hemi, omega

26:57

and Celi looks like, not to be confused with hemi atrophy.

27:00

So when you look at these cases, always is important to kind

27:04

of think about which side is abnormal here.

27:06

So in this case, the right side is abnormal.

27:09

Uh, we have actually enlargement of ifs lateral, uh,

27:13

uh, ventricle here.

27:14

So, um, this is a type of, uh, this, a example

27:18

of a Hemi Omega cephalic encephalopathy.

27:22

So al heterotopia, basically, you can have a rest

27:25

of gray matter, uh, migration anywhere

27:28

between the ventricle and the cortex.

27:30

So when this occurs in the, uh,

27:32

ventricle sub al heterotopia here, we have extensive kind

27:35

of angulating abnormal, uh, gray matter.

27:37

So these should follow, uh,

27:39

gray matter signal all sequences.

27:41

So on the T two, it should look just like the cortex.

27:44

And then pure rage, same intensity as the cortex

27:48

and flare, uh, usually should not have hyperintense signal.

27:53

This is a case of a band heterotopia.

27:55

So these neuro made it out a little further

27:58

that got arrested, kind of in the deep white manner.

28:05

And this is an example of a closed lip schizo celi.

28:08

Uh, these are always lined with abnormal poly MicroAge area.

28:12

So the cleft is, uh, lined by poly MicroAge area.

28:19

Let's see, 30 minutes.

28:22

Um, so, uh, now we're gonna kind

28:26

of focus more shifting gears toward the tumors.

28:29

Uh, going from benign to, uh, more high grade.

28:32

Of course, you know, higher grade, like grade three

28:35

and grade four tumors can cause seizures.

28:37

Uh, but we don't typically, you know, they're not epilepsy,

28:40

uh, causing tumors

28:41

because, uh, you know, they're their own entity

28:44

and are treated more aggressively.

28:46

So in terms of the, uh, more lower grade tumors,

28:49

we can talk about hematomas, which are basically arrests

28:53

or basically normal tissue and an abnormal place.

28:56

Uh, there's this new entity called multinodular evacuating

28:59

neuronal tumor, which I'll show, um,

29:02

which can cause seizures.

29:04

Uh, the classic kind of, uh, lower grade tumors

29:07

that cause seizures would be the dnet glioma, uh, PXA

29:12

and, uh, oligo dro gliomas are also common type

29:16

of tumors that can cause seizures.

29:19

So this multinodular evacuating neuronal tumor is, um,

29:23

is a recently recognized entity.

29:25

Um, pathology still don't agree,

29:27

whether it's a cy architectural pattern versus

29:30

a benign tumor.

29:32

What it looks like basically is like if you splash paint,

29:35

you know, in a subcortical white matter,

29:37

and it just, it always looks like this.

29:39

So almost like looks like prominent perivascular spaces, uh,

29:43

but they're more focal.

29:45

So they occur in the subcortical white matter.

29:47

They typically have no enhancement or mass effect.

29:50

Uh, they can be an incidental finding without any symptoms,

29:54

or they can be associated with seizures.

29:58

Hematomas, as I said before, are basically normal tissue

30:02

and abnormal place, so they're overgrowth of tissue.

30:06

Um, the classic one for boards

30:08

that we all study is the tumor cerium hematoma.

30:11

Uh, the tumor cerium is kind of like the floor

30:14

of the third ventricle, part of the hypothalamus,

30:16

but you can get hypothalamic gliomas

30:19

or, um, kind of anywhere.

30:22

Uh, they can be sessile or pedunculated.

30:24

The tumor scenario one is associated with seizures.

30:28

Uh, they should follow kind of gray matter on all sequences.

30:31

Sometimes they can have a little bit of gliosis,

30:33

so they may look a little bright on flare,

30:35

but mostly they should follow gray matter on the, uh,

30:39

all sequences, and they don't have any

30:40

enhancement or calcification.

30:44

Uh, it's important to kind of, uh,

30:47

differentiate these from hypothalamic gliomas,

30:50

which are actual tumors,

30:52

even though they might be low grade, uh,

30:54

they have neoplastic potential.

30:55

So the gliomas, they tend to be more T two hyperintense.

30:59

Uh, they tend to be more expansive.

31:01

Here we have one, uh, kind

31:03

of here on the T two weighted image.

31:05

You can almost, you know, miss it

31:07

because it's so, T two hyperintense almost blending

31:09

into the rest of the ventricle.

31:11

It's actually on the kind of fiesta, uh, case sequences.

31:14

You can really visualize, uh, this tumor in its mass effect.

31:19

These may or may not have enhancement.

31:21

Uh, typically in kids, there's going to, these are gonna be,

31:24

uh, pilocytic, astrocytomas and associated with NF one.

31:30

There's another example of, uh, emmic,

31:33

oh, this is actually the same case.

31:34

This is a pre T one.

31:36

Uh, here, post T one contrast, you see a little bit

31:39

of enhancement, uh, in here.

31:41

So having enhancement

31:42

or not, as you know, doesn't mean it's a high grade.

31:44

So pilocytic astrocytoma typically have enhancement in

31:48

their, uh, low grade tumor.

31:53

Uh, so now begin kind of to talk about the classic kind of

31:58

with the nodule type of tumors.

31:59

Uh, well, not yet.

32:01

This dnet is actually, uh, you know, typically, uh,

32:05

for board purposes, a bubbly corly based lesion.

32:08

Uh, it's, uh, type one,

32:09

it's a grade one glio neuronal tumor has this kind

32:12

of classic bubbly appearance.

32:14

So almost like a bunch of, uh, soap bubbles.

32:16

Uh, they are more commonly found in the temporal lobe, less

32:20

so in the frontal lobe.

32:21

They rarely enhance

32:22

or calcify, which is important to recognize.

32:25

Uh, when you kind of compare it to the other, uh,

32:28

things I'll show later, uh,

32:30

because they're very slow growing, if they're kind

32:33

of near the cortex,

32:34

they can actually scout the adjacent calvarium over time.

32:38

So important to kind of look at a ct if you have CT

32:41

or even on the MRI to see if you have that finding.

32:44

And that that's a feature that you'll see in a lot

32:46

of slow growing, uh, lesions.

32:48

But you can see that in dnet gang gliomas are,

32:53

uh, also type one glio neuronal tumors.

32:56

Uh, these are the classic kinda assist

32:59

with the nodule, uh, tumors.

33:01

Um, they have variable enhancement they tend to enhance, uh,

33:06

a lot more often than DAD nets don't, uh, commonly enhance.

33:10

And you tend to have, uh, more classification

33:12

with gang glioma.

33:16

Um, pleomorphic ex santo astrocytoma

33:19

or PXA are slightly higher grade.

33:21

They're grade two astrocytomas.

33:23

Uh, they also have this cyst, uh,

33:26

with enhancing nodule uh, appearance.

33:29

They tend to have a little more edema than the gang gliomas.

33:32

And, uh, they can even have a little bit

33:34

of a lepto menal enhancement.

33:36

Sometimes you'll see a little bit of, uh,

33:39

enhancement in the, uh, overlying lepto meninges.

33:45

This is not a, uh, interactive session,

33:47

so I can't have you guys answer questions.

33:49

So, th this is a caven smell formation.

33:51

So you can see this kind of, uh, susceptibility artifact,

33:54

T two, uh, dark, uh, lesion here

33:57

with some underlying white matter abnormality in some kind

34:01

of, uh, peripheral kind of sinusoidal kind of enhancement.

34:05

So this is a cavernous malformation.

34:11

So this is, uh, uh, a lesion in the left frontal lobe.

34:15

I guess you can feel free to put in the text what you think.

34:19

This is tumor or cortical dysplasia.

34:22

Um, but this turned out to be a tumor.

34:25

Uh, so this is a low grade tumor because it's expansile.

34:30

Um, we, uh, there was no significant enhancement,

34:33

um, and we did Mr.

34:35

Or Mr. Profusion and showed

34:36

that this was a hypo, uh, perfusion.

34:39

So, low grade tumors, uh,

34:41

typically don't have elevated perfusion.

34:44

So, uh, it doesn't, having kind of normal

34:47

or low perfusion doesn't exclude low grade tumor.

34:50

So this tend to, this, uh, turned out

34:52

to be oligo dro glioma.

34:55

Sometimes we do spectroscopy.

34:58

Uh, they try to differentiate between low-grade tumor

35:01

and cortical dysplasia.

35:03

And, uh, a normal MRS spectrum looks kind of like this.

35:06

Um, you know, this one didn't have the traditional hunter

35:09

peak, where you have the choline creatine,

35:10

and NNA, um, we see this sometimes, um,

35:14

doesn't mean it's abnormal.

35:16

Here we have an example of a focal cortical dysplasia.

35:21

Now, the problem with, uh, MRS is that everything causes,

35:26

um, increased choline decrease in a, so in this case,

35:29

the focal dysplasia, cortical dysplasia increased choline

35:33

because of, uh, increased membrane turnover, uh,

35:35

decreased NAA,

35:37

but it does that in a much lesser degree than the tumor.

35:41

So in the tumor, you have a greater increase in choline

35:45

and a much greater decrease in n na.

35:48

So, uh, papers have shown, if you look at the, uh,

35:50

the degree of, uh, the,

35:54

not just the direction of change,

35:56

but the actual magnitude of the change, you can kind

35:59

of differentiate between low grade tumors

36:01

and focal cortical dysplasias.

36:03

In practice, though, I think, uh, most people

36:06

who look at MMRS, uh, tend to agree

36:09

that these are not clear cut.

36:11

So there's no, um, universally accepted thresholds

36:15

for separating these things.

36:17

So it's kind of, you have to look at the lesion as a whole

36:20

and use your clinical

36:21

judgment checking on time.

36:27

Um, so we have, uh, 20 minutes left.

36:29

I'll go over just a couple of kind of syndromes

36:32

that you may see in adult neuro imaging.

36:35

Um, there's a bunch of pediatric epilepsy syndromes, which,

36:39

you know, I won't talk about today.

36:41

And then I'll focus a little bit on cephaloceles,

36:43

which are increasingly recognized as causes

36:46

of seizure or epilepsy.

36:49

So this is a case of a Sturge Weber syndrome.

36:52

Uh, you can see kind of calcifications.

36:55

Uh, so these, these are tend to to be childhood.

36:57

They're recognized in childhood, so usually

37:00

by the time they're adults, they, they should know

37:01

that they have s webber.

37:03

But you may see this, um, uh, basically calcifications.

37:06

You have, uh, lepto meningeal enhancement, um, due to the,

37:10

uh, abnormal kind of basically venous, uh,

37:13

drainage in a po angio mitosis that occurs in, uh, s Weber.

37:17

And they usually will have some kind of port wine stain

37:20

or, uh, facial, um, kind of, uh, uh,

37:24

stigmata on the ipsilateral side.

37:29

This is an example of a d**e Davidson Meison syndrome.

37:33

Basically, it's a cerebral hemi atrophy,

37:36

and this is more description of a constellation

37:39

of findings in the clinical syndrome rather than an entity.

37:42

So, uh, various things can cause this, uh, can be from kind

37:46

of a early insult, uh, some kind of encephalitis or,

37:50

or idiopathic.

37:51

So what you see is basically, uh, atrophy, uh, of one side

37:56

of the brain, as we can see here.

37:58

Um, surprisingly,

38:00

there's not much gliosis given the degree of atrophy.

38:02

So you know that the insult probably occurred very early.

38:05

Uh, you may see kind of thickening

38:07

of the underlying calvarium.

38:08

So here, uh, in kind of indicating that this is a kind

38:12

of a slow growing, slow, uh, process

38:15

that the calver actually thickened to fill that empty space.

38:18

And you can even see kind of, in this case,

38:20

we don't have it, but you can see kind of enlargement

38:23

of ipsilateral, uh, perinasal sinuses.

38:26

And here we have kind of more ln degeneration

38:29

of the ipsilateral ce.

38:31

Uh, this is a little oblique,

38:33

but you can also see kind of just the, the middle

38:36

of cranial fossa to be, uh, smaller on that side.

38:39

So th this is a clinical syndrome

38:41

that can be from various etiologies

38:43

or, um, the idiopathic, um,

38:50

switching gears to encephalocele.

38:52

So, uh, ence seals are basically protrusions

38:57

of brain through, uh, a defect in the, uh, in the calvarium.

39:02

So, uh, it can be a natural defect such as the FMO Valley.

39:06

So in patients with idiopathic intracranial hypertension,

39:10

because of the chronic elevated intracranial pressure,

39:13

you can actually have kind of almost a erosion

39:15

of the FMO valley here demonstrated on the

39:17

3D reconstruction.

39:19

So it gets larger and part

39:21

of the basal temporal lobe actually herniates through.

39:24

And sometimes on flare

39:25

or, you know, DIR,

39:26

you can actually see gliosis in that portion of the brain.

39:30

So now, in these patients, uh,

39:32

de encerios can be a seizure focus.

39:35

So they can occur, uh, in multiple different locations,

39:39

in this case, through a natural foramen.

39:42

But other times you can have these kind

39:44

of thoro osseous defects also as sequela of ih, uh,

39:48

basically have a, almost like a chronic pressure erosion

39:51

of the inner table of the area.

39:54

And you have this herniation

39:56

of the temporal pole into that area.

39:58

And you can even see there's xva dilatation of that, uh,

40:02

CSF space there, indicating that there's almost kind

40:05

of a tenting or, uh, traction of this temporal pole.

40:09

It's not just like, you know, freely going in there,

40:12

it's actually being pulled or stuck in there.

40:14

And this, uh, it was, uh, studied with a stereotactic EEG

40:18

and found to be the seizure source.

40:20

So, uh, these, uh, occur very commonly where I practice,

40:25

where there, you know, there's a lot of obesity, uh, a lot

40:28

of patients with ih.

40:30

Uh, and you know, up until a few years ago,

40:33

these were under recognized.

40:34

And, uh, you know, the patients were just, uh, you know,

40:39

thought to have just general, just un nonfocal epilepsy

40:43

or, uh, without a structural cause.

40:45

But they actually have this cause

40:47

and then they can be, uh, studied.

40:49

And these, um, areas can either be repaired

40:52

or actually, um, one

40:54

of our neurosurgeons does a laser ablation

40:56

of these, uh, areas.

40:58

So, uh, sometimes though, uh, patients with IH

41:02

commonly will have bilateral encephalocele,

41:04

so it can actually be sometimes difficult to determine which

41:08

encephalocele, uh, these are from.

41:11

So, uh, they won't be studied usually

41:13

with a stereotactic EEG to confirm.

41:17

Uh, ence seals can also occur elsewhere.

41:20

Uh, I've seen them, you know, in weird places,

41:22

even transverse sinus.

41:23

You can have herniation

41:24

of the brain into the transverse sinus.

41:26

In this case, uh, it's a frontal sinus, uh, ence seal.

41:31

Uh, this patient was, uh, actually admitted as an inpatient

41:35

with, uh, a meningitis and had this brain MRI,

41:39

and it looks initially on first glance

41:41

to be, you know, sinusitis.

41:43

But if you look closer, there's actually a discontinuity

41:46

of inner table of the calvarium.

41:48

On the coronal.

41:50

You actually see this gyro form

41:51

or cerebral form appearance of that tissue.

41:54

It's not just, uh, sinus disease.

41:56

And on the CTA, we actually saw vessels going into it.

42:00

So, which really cinched the deal for us

42:02

that this was an encephalocele.

42:04

And this patient actually has epilepsy, um,

42:07

from chronic meningitis.

42:09

They, they, they admitted multiple times in the past

42:12

with meningitis and had imaging,

42:14

and it was always, you know, called sinusitis.

42:17

And that was under recognized to be encephalocele.

42:23

So, uh, in conclusion, imaging is important

42:25

to detect treatable causes of epilepsy.

42:27

That's very important for, uh, treatment, uh, triage,

42:32

as well as kind of predicting the outcome.

42:35

Uh, the imaging finding can be very subtle.

42:38

So it's important to use kind

42:39

of high field MRI optimized techniques, uh, especially

42:43

to detect, uh, mesial temporal sclerosis

42:45

and focal cortical dysplasias.

42:47

And also, it's important to, you know, at least, you know,

42:50

attend seizure conference

42:51

and see how, um, the, the team, the epilepsy team kind

42:56

of uses all the information, uh, with PET and semiology

43:00

and, uh, the monitoring, uh, to come up the diagnosis.

43:04

Encephalocele are, uh, pretty common cause of FFC,

43:08

at least in some parts of the world where IH is more common.

43:11

And, uh, it's important to recognize those, uh, as well.

43:17

So with that, I'd like to thank you for your attention.

43:19

Uh, here's my email address on the left,

43:22

and you can also contact me via Twitter.

43:24

Um, we are having an open house

43:27

for our neuroradiology fellowship at, at end of the month.

43:29

So, uh, everyone who's, uh, a resident is welcome to join

43:33

that and learn more about, uh, neuroradiology at Emory.

43:37

And with that, uh, I'll start to answer some questions.

43:42

Okay, so someone asked no routine a SL to look

43:46

for Interictal Hyperperfusion.

43:48

Ictal hyperperfusion.

43:51

Okay, so, um, we don't typically do a SL, uh,

43:55

for epilepsy protocols.

43:57

Um, we usually don't image, uh,

44:01

with MRI in the ictal stage.

44:03

So, um, typically we're imaging in the interictal stage.

44:08

There are papers that say that in the interictal stage, uh,

44:12

you can use a SL almost like a pet,

44:14

you'll see, uh, hypoperfusion.

44:16

But, um, that is not part of our protocol.

44:19

Uh, most of our patients do get, uh, pet imaging as well.

44:22

So, uh, that that is true, that, uh,

44:25

you can use a SL in the same way,

44:27

but I don't think it's commonly used.

44:33

When do you refer patients for PET ct?

44:37

Um, so, uh, in our con in,

44:41

I guess the practice will be different, uh,

44:43

depending on where you practice.

44:45

At our center, um, most patients will get a pet.

44:48

Uh, other places, you know, you may have, um, patients who,

44:52

you know, who have discordant imaging findings

44:55

with their clinical feature or some kind of

44:57

more atypical cases, they'll get pet.

44:59

But at our center, um, it's part

45:01

of our standard epilepsy protocol for refractive epilepsy.

45:05

They'll get PET imaging. Um, as part of that workup.

45:13

Another question, isn't enhancing gliomas

45:17

or aren't enhancing gliomas as malignant?

45:20

Um, not necessarily.

45:21

So in the adult population, uh,

45:26

I guess you can phrase it differently.

45:28

Um, so non enhancing tumors

45:33

are typically lower grade,

45:35

but having enhancement doesn't

45:37

really mean it's higher grade.

45:39

Um, so typically when you see a non enhancing glioma,

45:42

typically you're talking about low grade

45:44

glioma, but that's not always true.

45:45

You can't have, you know, um, you know, types

45:48

or WHO uh, grade three tumors that don't enhance

45:53

or have minimal enhancement.

45:56

But, uh, some tumors like pilocytic astrocytomas

46:00

or px, a, even like gliomas, even D nets,

46:03

they can have enhancement and they're not malignant.

46:11

So click here, click here

46:14

in Ictal PET or Interictal pet.

46:17

So typically, uh, PET is done interictal,

46:22

uh, because just the, you know, the logistics of, uh,

46:25

doing the PET with, uh, preparing the FDG

46:28

and everything, um, we do sometimes do spec, uh,

46:33

with ictal and interictal spect

46:35

and obtain the subtraction image.

46:38

Um, so we typically do ictal spect, but not ictal pet.

46:45

Someone asked, when is MEG useful?

46:48

Uh, MEG uh, is also an adjunct to,

46:52

so it can be part of a comprehensive epilepsy workup.

46:56

Uh, MEG is not always available.

46:58

It's actually, you know, only available at a few centers.

47:01

So when there is, uh, kind of atypical findings, uh,

47:05

sometimes, uh, we will refer patients for MEEG

47:09

and then, you know, uh, so it is a discussion

47:12

that usually occurs in the epilepsy conference.

47:16

And if they feel confident that you know about a, uh,

47:19

seizure onset zone, they may go after it.

47:22

You know, if they don't feel confident,

47:24

they may need more information obtained MEG

47:26

or additional, uh, invasive monitoring even

47:30

before they, uh, decide the treatment course.

47:35

Is it required to refer

47:37

to three T when 1.5 T shows no abnormal,

47:41

no abnormality in the epilepsy patient,

47:44

or can be treated like idiopathic?

47:46

Um, I think nowadays

47:49

three T probably is the standard of care.

47:52

I think most centers have three T

47:54

and it does increase their sensitivity for subtle, uh,

47:58

pathologies, especially MTS, um, and SCD,

48:02

and that, that makes a difference, uh, for treatment.

48:05

So, um, typically it's okay to get one point 5g,

48:09

like if you have an inpatient, uh, you know, er patient, uh,

48:13

presenting with seizures, you know.

48:16

But, uh, if a patient with a chronic refractory epilepsy,

48:20

at some point they should get a three T just

48:22

to increase the sensitivity of detecting a,

48:25

uh, structural cause.

48:30

What is experience using A-S-L-A-S-L do you recommend?

48:33

We don't, uh, I think I answered this question earlier.

48:36

We don't typically use a SL.

48:37

It's kind of a, uh, there are some issues, you know,

48:40

with standardization and, you know, in terms of that,

48:43

but, you know, some papers have shown that, you know,

48:46

you do have, uh, hypoperfusion, uh,

48:49

in the interictal period can, um, kind

48:52

of give the similar kind of information to a pet.

48:54

But we don't do that clinically here, if not standard

48:59

to do a PET cd.

49:02

What are the indications?

49:03

So I kind of answered that earlier as well.

49:07

That depends on your institutional practice.

49:09

At our institution, usually with refractory epilepsy,

49:13

we typically do get a pet, so it is part of our standard,

49:17

uh, other places, um, they may do it if, for example,

49:22

the SEMIOLOGY and the MRI, uh, does not match,

49:26

or the MI doesn't reveal much, um,

49:29

or you have a little bit of atrophy of the hippocampus,

49:31

you don't know whether it's just natural asymmetry

49:34

or it's actually MTS.

49:35

So in those cases during the pet, you know, we give you kind

49:39

of corroborative information.

49:43

What about PET MRI,

49:45

let just check the time here that we do have time.

49:48

So, uh, PET petm, I, uh, interesting question.

49:51

Um, you know, there's a lot of interest with petm.

49:54

I, and they're kind of different vendors now.

49:57

Uh, I think major vendors all have clinical, um,

50:02

clinically approved devices,

50:04

and we are actually, we don't currently have one,

50:06

but we're getting one installed, um,

50:08

probably in within the next year.

50:10

Uh, the advantage of PET MRI in neuro imaging is less

50:14

than in body imaging, uh,

50:16

because registration for the brain is typically easier.

50:20

So, um, we can almost always, uh, obtain, you know,

50:24

pretty accurate registration, uh, for, uh, a PET CT

50:29

and A MRI, uh, the only kind of convenience in terms

50:33

of only advantage for brain imaging for PET MRI is

50:36

that you can get simultaneous, uh, PET

50:39

and MRI, which is, uh, convenient for the patient.

50:42

Uh, but it's also, you know, important for research, uh,

50:46

if you're actually testing different kind of radionuclides,

50:49

not just FDG, uh, having pet MI, it's, it's really helpful

50:54

to kind of validate your different kind of, uh,

50:56

radionuclides with, uh, your different agents with, uh,

51:00

say like functional MRI or a SL

51:07

poly micro jia features.

51:11

Um, poly micro JIA has many forms.

51:13

I'm not a pediatric neuroradiologist, so, um, don't pretend

51:17

to be an expert in poly micro jia,

51:19

but t typically, um, kind of, uh, it looks like, um,

51:25

just, uh, instead of the normal jro phone, uh, you,

51:28

you have multiple, just a finer pattern of, uh, gyration.

51:33

So, uh, you can, it, it basically, um,

51:38

it looks more bumpy as opposed to having a smooth cortex.

51:41

Uh, it's hard to kind of, uh, explain on, uh,

51:47

using words, but, um, you know, it's, it's typically kind

51:50

of a, almost like a cobblestone kind of appearance

51:53

rather than just a smooth, uh, Jarrow pattern.

51:59

What should be sequence of interpreting epilepsy protocol?

52:02

MRI? Um, so the sequence of interpreting epilepsy

52:08

protocol MRI probably will differ based on your

52:13

local, um, MRI protocol.

52:17

So depending on what, uh, protocols obtain, typically

52:21

what I do is I just look at all the sequences once,

52:26

and then I focus on the coronal T two and flare.

52:31

Um, just

52:33

because MTS is such a common cause of epilepsy in adults,

52:37

so carefully look at the, uh, coronal and flare,

52:41

and not just at the hippocampus, but also at the amygdala

52:45

and all those limbic structures that's associated.

52:48

And also for encephalocele along the, you know,

52:51

middle cranial fossa.

52:53

Then I will go to the MP rage

52:55

and closely look at, you know, choose, you know, do

52:58

that in any pattern you want.

52:59

And closely look at the gray white differentiation of the,

53:02

uh, different ri

53:08

kindly suggests dedicated epilepsy research institution

53:13

to refer patients.

53:16

Um, not sure what, what, uh, the, uh,

53:21

attendee means by epilepsy research

53:24

institution institution.

53:27

Uh, I guess, I mean, depending on your, uh,

53:30

where you're close to, they're, you know, usually most large

53:35

academic centers have a big epilepsy program.

53:38

Uh, we have a pretty big epilepsy program,

53:39

but so do a lot of academic centers.

53:43

I think, um, you know, working with your neurologist

53:47

and, um, your, you know, team at your institution,

53:51

it's probably best to help you answer that question.

53:57

Uh, most of the images on 1.5

54:00

or three T, so most

54:03

of the images are, uh, at three T.

54:06

So we typically, um, uh,

54:10

protocol our MRIs when it's done

54:12

for refractive epilepsy at three T,

54:14

they're actually usually done on one scanner.

54:16

Uh, we, we like to kind of, uh, do a function MRI

54:19

as well at the same time,

54:20

and especially if they're considered for epilepsy surgery.

54:23

So we want to know that language dominance, um, and such.

54:27

So, um, we'll do them on this dedicated, uh, three t uh,

54:32

scanner with, uh, also usually same day.

54:35

FMI thank you for very informed lecture.

54:41

Oh, hold on.

54:44

Which sequence of m MRI is important?

54:48

Which sequences of MRI are important for, uh,

54:51

looking at the cause of epilepsy?

54:54

I mean, I think you, you have to look at all the sequences,

54:56

but you know, um, now the, the major, the major ones I,

55:01

you know, are dedicated for

55:04

epilepsy protocol would be the high resolution NP rage, uh,

55:07

which you're looking at the gyration pattern,

55:10

the gyro pattern differentiation,

55:14

but also the, uh, oblique coronal T two

55:16

and flare, where you're really looking closely at

55:19

the medial temporal lobes.

55:23

When do we call idiopathic epilepsy?

55:26

So that's really, um, not a call that, you know,

55:29

radiologists would make, uh, typically when a epilepsy

55:34

team kind of, um, goes

55:37

through their whole comprehensive workup

55:39

and still don't come up with a cause, um, that that's,

55:43

uh, considered idiopathic.

55:45

But you know, that's all relative, right?

55:48

So idiopathic epilepsy,

55:49

you don't find anything at three T

55:51

70, you might find something.

55:53

So that, that's a kind of a subjective, um, kind

55:57

of bucket list, a bucket term.

56:02

Do you have a resting state? FMRI, uh, pipeline?

56:07

Uh, we do have a FMI pipeline, uh, for resting state.

56:12

Um, the problem with resting state FMI is

56:15

that there's no FDA approved, uh, analysis software.

56:19

So, you know, you can't, uh, just rely on resting state.

56:22

So we, so we always do task-based FMRI, uh,

56:25

for language lateralization, um, resting state, FMI,

56:29

you know, it's pretty reliable

56:31

for localizing sensory modal cortex, uh, visual cortex,

56:34

those things for language, uh, we find

56:37

that it's not really very lateralizing.

56:40

Uh, perhaps it's actually showing more of a network

56:43

for a language network, but it doesn't really, um,

56:46

consistently show which side is more, uh,

56:50

I guess more important

56:51

or more, uh, important is not the right word.

56:55

That would, which side would suffer the most if we resected

56:59

lead to more language deficit.

57:01

So typically it's less lateralizing than test-based.

57:03

FMI, I think, uh,

57:10

idiopathic epilepsy is the most common,

57:13

or MTS, um, I would say

57:18

that also depends on, you know, how, uh,

57:22

comprehensive evaluation is,

57:24

because idiopathic is kind of a, uh,

57:28

ill-defined kind of a term,

57:30

but, um, you know, usually, uh, about 50

57:35

to 60% of cases are idiopathic

57:38

or don't have a, a known cause.

57:41

So by that, you know, you can kind of guess

57:44

that is probably more common than MTS

57:47

because there are other things that can also cause epilepsy.

57:49

So I guess idiopathic is more common than MTS.

57:59

Okay. I think, uh, our time's up

58:01

and, uh, I think I've answered all the questions.

58:03

Again, feel free to email me if you have any questions.

58:06

Uh, contact me

58:07

and then, uh, if you're a trainee, uh, welcome

58:10

to attend our open house on the 24th. Thank you very much.

58:15

As we bring this to a close, I wanna thank Dr.

58:17

Who for this amazing lecture.

58:18

And thanks to all you guys participating in our no

58:20

conference or minor.

58:22

This conference is available on demand on MI online.com.

58:25

In addition to all the previous new conferences, be sure

58:28

to join us tomorrow for a lecture from Dr. Bruce Forrester

58:30

on Synovial Linings.

58:32

You can register for that@mrionline.com

58:34

and follow us on social media at the MRI online for updates

58:37

and reminders on upcoming new conferences.

58:39

Thanks again everyone, and have a great day.

58:43

Thank you.