Lea Alhilali, MD Profile picture
Sep 2, 2022 15 tweets 9 min read Read on X
1/Feeling lucky? Or feeling evidence based?
A #tweetorial about how to read a #pituitary #MRI using data and know if there’s cavernous sinus invasion w/the Knosp score.
#medtwitter #FOAMed #radres #neurorad #FOAMrad #neurosurgery #medicalstudent #meded #USMLE #endocrinology Image
2/The hardest part of a pituitary MRI is deciding if there is cavernous sinus invasion. It makes sense that the more lateral a tumor goes on MRI, the more likely it has invaded the sinus—bc it is going the direction of the sinus. But how far is far enough? Image
3/This is important bc each time a radiologist makes a call on imaging, they make a bet & they are betting their credibility. And unlike other bets, there is only 1 wager—all in! So it is important to not call it when you might be wrong, bc overcalls destroy credibility. Image
4/We know for medial tumors we shouldn’t call it bc the chance of being wrong outweighs being right. And if it is very lateral, chance of being right outweighs being wrong & we should call it. But where does the chance of being right outweigh being wrong? Knosp score will tell us Image
5/Knosp score is based on the position of the tumor w/respect to the ICA. Cavernous/supraclinoid ICA is shaped like a macaroni elbow, so when you cut it in cross section, you see circles that are the two ends of the macaroni—top one is supraclinoid ICA, bottom is cavernous ICA. Image
6/Knosp grade draws 3 lines along the circles—(1)medial carotid line, medial to the circles, (2)intercarotid line through the center of the circles & (3)lateral carotid line, lateral to the circles. They are like 3 traffic lights—w/the color=the chance of cav sinus invasion Image
7/At the most medial the tumor does even reach the medial line. There is essentially no chance of cav sinus invasion (7%). But you don’t need a Knosp score for this—a tumor that doesn’t really approach the cav sinus probably doesn’t invade it (thank you Captain Obvious!) Image
8/As the tumor goes more lateral, it crosses the medial carotid line. But it only crosses the green line, so green is still good to keep moving & pass up calling invasion. Only 1 in 5 of these will have invasion, so if you call it, you will be wrong 80%--not good Image
9/Going more laterally, now you cross the intercarotid line—the yellow line. Yellow means caution or slow down. These will invade the cav sinus in about half of cases. So it is enough to make you slow down, and take a good look, but not enough to stop and call it. Image
10/This is because 50% is still essentially a coin flip. You wouldn’t put all your money on black, so you shouldn’t put all your credibility on something that could be wrong half the time. So you want to hold back your chips here and say that it does not definitively invade. Image
11/Finally you cross the lateral carotid line—this is the red line—which means stop and call it. This is bc you will be right almost 90% of the time. Red = bad = stop = in the cav sinus Image
12/This is a bet that you want to make—you would put serious money on a bet that has a 90% chance of winning (you probably put it on bets that are less in the stock market nowadays!). So take advantage of knowing you will be right and call it. Image
13/The last Knosp grade is when tumor surrounds the cavernous ICA. This has a 100% chance of cav sinus invasion. No surprise that if the tumor completely surrounds something in the cav sinus that the tumor is also, wait for it…in the cav sinus. Thanks Captain Obvious. Image
14/To summarize, think of the Knosp lines as 3 traffic lights—the color of the line crossed tells you what you should do. Cross medial green, keep going—there is nothing to call. Cross middle yellow—slow down, it is close but not definitive. Cross lateral red, stop & go all in. Image
15/So now you know the Knosp score & how it can help you to be right. So use it and don’t settle for a cheap Knosp off 😂. Let the data be your superpower! Image

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Lea Alhilali, MD

Lea Alhilali, MD Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @teachplaygrub

Nov 20
1/Time to rupture all your misconceptions about aneurysms!

When you see an aneurysm on imaging, do you know if it’s at high risk of rupture?

This month’s @theAJNR SCANtastic shows you which aneurysms are bursting w/risk!

ajnr.org/content/45/11/…Image
2/Aneurysm rupture is a devastating even, as it results in subarachnoid hemorrhage & complications such as hydrocephalus, vasospasm, infarcts, & death.

Preventing it by treating aneurysms before they rupture is key. But you also don’t want to overtreat. Image
3/To remember what features make an aneurysm more likely to rupture, think what makes that guy at the bar that you angered more likely to rupture & start a fight.

What makes him more likely to rupture are the same things that make aneurysms more likely to rupture Image
Read 19 tweets
Nov 11
1/Need help reading spine imaging? I’ve got your back!

It’s as easy as ABC!

A thread about an easy mnemonic you can use on every single spine study you see to increase your speed & make sure you never miss a thing! Image
2/A is for alignment

Look for:
(1) Unstable injuries

(2) Malalignment that causes early degenerative change. Abnormal motion causes spinal elements to abnormally move against each other, like grinding teeth wears down teeth—this wears down the spine Image
3/B is for bones.

On CT, the most important thing to look for w/bones is fractures. You may see focal bony lesions, but you may not

On MR, it is the opposite—you can see marrow lesions easily but you may or may not see edema associated w/fractures if the fracture is subtle Image
Read 11 tweets
Nov 8
1/Raise your hand if you’re confused by the BRACHIAL PLEXUS!

I could never seem to remember or understand it—but now I do & I’ll show you how!

A thread so you will never fear brachial plexus anatomy again! Image
2/Everyone has a mnemonic to remember brachial plexus anatomy.

I’m a radiologist, so I remember one about Rad Techs.

But just remembering the names & their order isn’t enough.

That is just the starting point--let’s really understand it Image
3/From the mnemonic, we start with the roots—the cervical nerve roots.

I remember which roots make up the brachial plexus by remembering that it supplies the hand.

You have 5 fingers on your hand so we start with C5 & we take 5 nerve roots (C5-T1). Image
Read 20 tweets
Nov 6
1/Asking “How old are you?” can be dicey—both in real life & on MRI! Do you know how to tell the age of blood on MRI?

Here’s a thread on how to date blood on MRI so that the next time you see a hemorrhage, your guess on when it happened will always be in the right vein! Image
2/If you ask someone how to date blood on MRI, they’ll spit out a crazy mnemonic about babies that tells you what signal blood should be on T1 & T2 imaging by age.

But mnemonics are crutch—they help you memorize, but not understand. If you understand, you don’t need to memorizeImage
3/If you look at the mnemonic, you will notice one thing—the T1 signal is all you need to tell if blood is acute, subacute or chronic.

T2 signal will tell if it is early or late in each of those time periods—but that type of detail isn’t needed in real life

So let’s look at T1Image
Read 20 tweets
Oct 29
1/To call it or not to call it? That is the question!

Feeling wacky & wobbly when it comes to normal pressure hydrocephalus?

Don’t want to overcall it, but don’t want to miss it either!

Check out the latest in NPH w/this month’s @theAJNR SCANtastic!

ajnr.org/content/45/10/…Image
2/NPH was first described in 1965—but, of the original 6 pts, 4 were found to have underlying causes for hydrocephalus.

This begs the question—when do you stop looking & call it idiopathic? When do you suggest it on imaging? Image
3/There’s an iNPH Radscale, which scores 7 different imaging features.

Score above 8 is very sensitive for iNPH.

But who’s going to take out calipers & evaluate SEVEN different imaging findings on every dementia MR?

Also this scale doesn’t predict who will respond to shunting Image
Read 14 tweets
Oct 18
1/Do radiologists sound like they are speaking a different language when they talk about MRI?

T1 shortening what? T2 prolongation who?

Here’s a translation w/an introductory thread to MRI. Image
2/Let’s start w/T1—it is #1 after all! T1 is for anatomy

Since it’s anatomic, brain structures will reflect the same color as real life

So gray matter is gray on T1 & white matter is white on T1

So if you see an image where gray is gray & white is white—you know it’s a T1 Image
3/T1 is also for contrast

Contrast material helps us to see masses

Contrast can’t get into normal brain & spine bc of the blood brain barrier—but masses don’t have a blood brain barrier, so when you give contrast, masses will take it up & light up, making them easier to see. Image
Read 20 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(