Introduction to Clinical MRI Physics (part 3 of 3)

Intended audience: radiology residents and fellows, medical students, or anyone who is interested in learning basic MRI physics conceptually. No complex equations or formulas.
0:00 - Intro
1:00 - 2D vs. 3D
8:27 - Matrix and Field of view
11:00 - Wrap around artifact
12:05 - Nyquist sampling rate and Undersampling
14:04 - Frequency undersampling artifact
17:34 - Phase undersampling artifact
21:08 - Pulse diagrams and basic parameters
23:57 - Gradient echo technique
31:31 - Inversion recovery technique

[@HawtStyle3000]

I don't know how but you made everything simple and straightforward. I really appreciate this video series and really want to share my thanks.

[@matthewzebedee1465]

Thanks for these videos. They have explained in ~2 hours what months of reading textbooks have failed to do! Great simple explanation of matrix size vs FOV!

[@deepakv8735]

The best class i ever had. Id recommend this to everyone to understand the basics for a radiologist.

[@pafshin]

Thank you for these three videos. As a technologist learning about MRI physicists, I found out these series of videos are very helpful to understand the text books. I wished you could have made more video about MRA and MRV, follow compensation , chemical shifts, Dixon and sat bands. I understand you are a Radiologist but these video with illustrations are amazing.

[@faizfarhan8231]

Best explanation of MRI on youtube 👍

[@YukaTanMT]

I can finally understand MRI much better especially matrix. I love the illustrations. Simple to understand although there are some confusions/hallucinations but I’ll just have to rewatch your videos again and supplement with books.

[@harris4917]

Best MRI physics EVER. You should expand the topics and build on those videos as is the best attempt to explain a rather complex topic that stops thousands of doctors to understand better that modality.

[@dvillalobosv]

Thank you so much for your videos 🙏🏻 it helped understand all these topics that can be so difficult to understand. I appreciate your videos so much. I wish you had MRI contrast videos, or tips and tricks on how you memorized all of the MRI equitations and terminology! Again thank you so much for you videos !

[@applekim2006]

thank you so much for the great explanation.

[@coldmagnet]

Truly magnificent. The visuals, the illustrations really did an amazing job bringing me up to speed. I know its a gross oversimplification, but for exam purposes it doesn’t get better than this. Let me know if you plan to cover anymore topics like MRI artifacts, MR QC/QA, contrast agents, etc. Great stuff!!

[@neuroradish]

I'm glad to know it's helpful. Thank you. Perhaps down the road I'll do more MRI physics but most focuses will be on clinical, especially during the board exam season. I'm no physicist after all.

[@coldmagnet]

@@neuroradish Wouldn't want it any other way. Clinically oriented is much easier to grasp because it actually has a direct attachment to real life application.

[@hassanebrahim1131]

Stunning , thank you sir

[@naveenkumar_]

Simplified .. thank you very much.. helped a lot in understanding

[@kennethwilliams9353]

OMG The same Phase coding Techniques was used to study the ionosphere at Arecibo observatory which is no more unfortunately. You explain things very clearly honestly. I love your remarks are you still awake the fat in Iowa so hilarious you’re very funny and smart also a good teacher.

[@neuroradish]

Thanks. I didn't know about Arecibo (R.I.P.); I learn new things everyday. Very cool.

[@kennethwilliams9353]

Yeah there’s some Doctor Who acknowledges immediately that he is not a physicist but he is a doctor a medical doctor he had a three part online seminar for people I think that wanted to be MRI operators and so it’s like basically a class on MRI physics and he made a remark I forgot his name and you know how the fat turns white in the Biased number one or something along that line it was just so funny because I’ve been to Illinois and Iowa and beautiful people are really good people and when they go to a restaurant do you know they don’t eat they basically feed I told the manager because they came over to our table and asked how was everything I said the food is so good and everything but you know what you put too much food on the plate. And so my friend and I were looking at the pies in the window case in the showcase and we’re gonna have to buy a pie to take with us on our travels because they look so delicious but we’re full and you put too much food on the plate just slowly do you know every month or so put one less teaspoon of whatever it is or that are slices of that meat you put way too much food on the table that’s why these people are huge but they’re lovely people I love them all. So only a few people understood what he was talking about but yeah him yeah I barely remember that but it sure was funny yeah he explains things quite clearly very clearly and is it wow that that sounds like he’s bouncing radio waves off of the ionosphere you know it with his face coding and everything if you if you run an MRI or listen to his class you know he talks The different frequencies for the different tesla or magnetic field and now that I’ve studied this stuff and forgotten about most of it a 3T is the minimum tesla that one should use is actually 3.1 and the RF frequency I think it’s 128 or something like that but you know what I was looking for something else not to operator the MRI machine but you know it’s interesting if you’re into MRIs regular MRIs don’t really help a lot what you really want to do if you want to know if these problems with peoples spinal cord or there do you know earn spondylosis and you know the narrowing of the spinal canal and isn’t doing any harm to the spinal cord you really should take I had a fusion MRI because it shows all of the fluids flowing and the spinal cord because you’re gonna have stuff arthritis but I don’t stenosis any type of narrowing of this bulge whatever and it may look like wow this person is really in trouble but actually with the diffused MRI to see the health how healthy or unhealthy the spinal cord is so that’s the definitive test is it a fusion MRI and an Jeff Cantor. He’s out cutting edge doctor he uses ultrasonic scalpels and techniques that don’t tear down here spinous process and then I have to rebuild it a lot of doctors don’t even understand what a ultrasonic scalpel is but I’m just telling everybody I can about ultrasonic scalpels they don’t do any harm to the nervous system and if you look at his example is only three minutes long are you get the long version but it just shows you that it cuts through bone in arthritis it basically vaporize it but it will not harm The nervous system

[@kennethwilliams9353]

Well you know what when it’s not so late at night and I can give you some references do you know some hypertext to show you who to watch for the latest technology an MRI and spinal surgery I will get back to you I didn’t even added this thing and I hope that it’s somewhat helpful but I have to go back and look at my notes I was only looking at this because I have I need to know my back I need a operation but I refuse the 70 year old technology that this guy wanted to rip out all of my spine is processes and fuse L23 and four I mean that’s completely unnecessary because now I’ve studied this and I know this for it to be true I have spondylolisthesis which is a superior vertebrae slipping on her inferior vertebrae and the or from the posterior to the anterior direction. So I stop studying that because I know what I need now I just have to wait till somebody in Los Angeles or California can do this ultrasonic work or I’ll have to fly to Florida and get the work done there because it’s so horribly painful anyway

[@wajeehasaeed3211]

Excellent

[@DrRad-mp2xq]

Wow I am really grateful to you

[@attheajanelepiten5422]

Thank you for this! Very helpful! Can you upload more? :D

[@shujatali642]

Super, keep it up.

[@jerry5550]

Why taking a whole year to learn , while you can just watch this video lol

[@bradoliva383]

Thank you very much Dr. Wu! I was wondering for spin echo sequences I have often seen them diagramed with an additional frequency encoding gradient applied between the 90 degree and 180 degree pulses (in addition to at TE/readout), and I was wondering what the purpose of this was? I was assuming it was to correct inhomogeneity in the field and allow transverse magnetization to dephase if given with enough time prior to readout when it is applied again? Also these are wonderful videos, I would love to see an additional lecture solely focused on MRI artifacts from you as you make these difficult concepts very easy to understand!

[@neuroradish]

Thank you. I'm glad it's helpful. To answer your question. You're correct, that extra frequency encoding gradient before 180 degree pulse is the "dephase lobe". Since the frequency encoding (readout) gradient needs to stay on the entire time during sampling in order to create different frequencies for localization. Even though it is relatively short, having this gradient would also cause phase shift and therefore loss of signal. To correct that, we purposely "dephase" them first, so during the readout, they are now rephased again during the 2nd application of the frequency encoding (readout) gradient. So at TE, the net effect of these phase shifts would be basically zero.

[@bhupeshgoyal6019]

🤟👍

[@shakeebuddinkashif4578]

I have a doubt regarding 180 degree pulse pancake flipping, you are showing an image that is not changing the phase of 45 degree spin by 180 to make it 225 but instead to 360 minus 45 which is 315. Why? How is it a 180 degree pulse then

[@davidmusoke]

Hey, thanks again for another excellent video. I was relieved it was easier to understand than the excellent video 2. Now, I have this nagging question about slice, phase encoding, and frequency encoding gradients. WHO (or what machinery) generates these gradients? I know the MRI machine has the main powerful B0 coil for proton alignment and XYZ gradient coils for image generation and readout. But I see no coils or other generators for slice selection, frequency encoding, or phase encoding. Sorry for the newbie question...

[@neuroradish]

No worry. MRI physics is a tough subject, and they are all great questions. I had (still have) many questions as there are so much to learn. For each MRI, as you know, there is that all powerful, always-on main magnetic coil B0 which runs in the Z or longitudinal axis (head to toe). Then there are 3 sets of built in coils that can be turned on and off by an operator, built in the Z, X, and Y axis. Please see the diagram in Part 2 at 4:00 time stamp. These Z, Y, and Y coils are used to generate gradients for slice selection, frequency encoding, and phase encoding. For example, for slice selection, you are trying to excite only protons in a particular location - a thin slice (volume) of a part of the body - that would be in the Z-axis more commonly. So you great a magnetic gradient first, so that the protons in your target zone are precessing at different frequencies than the rest of the body. Otherwise if you don’t have a gradient on, all portions throughout the body would be precessing at the same frequency (roughly 64 Mhz @ 1.5T). Then when you send in an RF pulse at 64 Mhz, all protons will be excited because all will resonate at this matching frequency. So you turn on your Z axis coil to generate a gradient in the Z-axis, to make sure that now protons across the body are precessing at different frequencies based on where they are on the gradient. Because you know their unique frequency at any point along the Z-axis gradient, you can select which area you want to excite using a matching frequency to that target area. So you have just “selected” a slice. The phase and frequency encoding work the same way to generate their gradients. Just that you choose axis that are perpendicular from one other. BTW, you don't need to make slice selection in the Z-axis only. You can slice select in Y or Z or any other oblique planes (by turning on a combination of coils simultaneously to generate an angled gradient); you just need to do subsequent phase encoding and frequency encoding gradients in the perpendicular planes. I hope that clarify a little bit. Like I said, MRI physics is a tough one for us radiologists / non-physicists. It took me multiple multiple tries before some concepts start to sink in. Cheers,

[@davidmusoke]

@@neuroradish So, its the XYZ gradient coils responsible for the slice selection, frequency, and phase encoding. No extra hardware needed? Thanks!

[@neuroradish]

@@davidmusoke you got it, my friend..

[@davidmusoke]

@@neuroradish Thanks ... now I have a more general question and since you are the resident MRI expert online ... -) 1. If a proton at position x y, z in tissue experiences an x-gradient field Bx = 1, and likewise By = 2 and Bz = 3, with the main static field being Bo, is the combined magnetic field it experiences Bo + Bx + By + Bz = Bo + 1 + 2 + 3 = Bo + 6 or do the x,y,z fields combine in some other way? Let's assume Bo is in the z-direction. 2. All textbooks and notes on MRI I've read talk about tipping the net magnetic moment vector to angles up to 180 degrees maximum(e.g for FLAIR or fat suppression). So what happens if the external RF pulse is long enough to cause 10 full rotations of the vector (3,600 degrees) PLUS a 90-degree tip on the 11th rotation to perform a simple T2* operation. How long does the recovery take? Does it take T1 seconds, like in a regular recovery? Does it rotate in the transversal plane 10 times, then start its recovering on the 11th cycle (for a 10x longer T2* interval)? Or do you get 10 T1 recovery cycles? Very strange question, I know but I was curious about it :-)

[@neuroradish]

@@davidmusoke I think you have begun dived into physics questions that are beyond my clinical interests, to be honest. But the magnitude and direction of the gradient a particular proton experiences at a given location would depended on which gradient coil is turned on at that time. If for example, only the set of Z coils are turned on, then it would experience either (B0 + Bz) or (B0 - Bz), depending on if it is located left or right to the isocenter. If it's at the isocenter, then it's just B0. Remember, the strength of the magnetic field generated by these gradient coils alone (X, Y, or Z) is significantly weaker than main magnet B0. Also for example, you turn on X and Y coils at the same time with equal magnitude, then you'd get a 45 degree gradient between X and Y. If you flip the spin to more than 180 degree, it would be just going back toward the longitudinal direction. So for example, if you do a 210 degree flip, it would be essential same as doing a 150 degree flip. You won't get a longer T1 recovery time (than 180). In fact, you would shorten T1 recovery time compared to a 180 degree flip. So doing a 3600 degree "inversion" would not give you 10x T1 recovery times. You will, however, cause a huge amount of energy deposition in the specimen and probably cook the meat :)

[@DrVikasChauhanDM]

Eureka after so many years of struggle...

[@taman449]

But why fat will travel faster ?

[@neuroradish]

Hi, I assume the question is why fat has shorter T1 relaxation / faster T1 (than water)? I'm not an expert in chemistry, so I probably can't explain it too well, but it has to do with the molecular structure that fat's hydrogens are held in a long chain and packed together, different than the small H2O molecule. So fat actually moves much slower (slow molecular "tumbling" rate) than water, at a rate for efficient T1 relaxation. Therefore it has a much shorter T1 value than water. By the way, because the hydrogens are packed tight close together in fat, they also have a much more efficient spin-to-spin interaction than water, so fat also has a shorter T2 relaxation (vs. water has long T2).

[@josephdays07]

For to do the wavelet transform or Function Wave i am not requiere the complex number or Fourier series. Just I need the new methodoly I discovered.I left this video to compare: kzbin.info/www/bejne/aXbFp6ymn5pqbac