Livesaver, i could not for the love of god figure out the phase encoding gradient like ever. After this i realised we went over phase changes back in middle school and with the help of ur vids it completed the picture for me. Will definitely recommend u to everyone studying this. ❤
This content is just amazing. I have spent hours upon hours watching videos on other channels and reading on websites like MRI Questions and still felt unsettled and unable to integrate the various concepts. After watching this, I feel SO MUCH better and feel I now have a solid conceptual understanding (especially understanding how the Fourier transformations help in identifying unique spatial frequencies). I wish I watched this at the start as it would have saved me such an awful headache! Thank you so much!
Honestly comments like these are the greatest reward for making these videos, happy to hear it helped make things click. Credit to MRIQuestions.com and Dr. Elster for the underpinning theory here, I just expanded on it a little :) And while the real picture building process is a little more complicated, this ties in well with the typically taught base MRI physics concepts while avoiding the math. If you're really curious to how the real thing works, stay tuned! We will have a full series on it in the future including K-space.
What is PHASE? Imagine we are singing a song together and there are three different possible variations: 1) One of us is louder or quieter than the rest, that is a matter of Amplitude. 2) Amplitude of our voices are the same but one on us is singing faster or slower that the rest, that is a matter of Frequency. 3) Amplitude of our voices are the same (nobody is louder or quieter that the rest) and Frequency of singing are also the same (nobody is singing faster or slower the other people), but one of us is a little ahead or behind than the rest, that is a matter of PHASE (Time, but it is measured in angular degree, from zero to 360).
thank you so much for your work. To be able to explain this well you had to understand it so well yourself and then trace back the process to each logical step! impressive!
Before this moment 20:59 I was very interesting by your fantastic explanation and pretty animations, but suddenly you go through the nexts.. Unfortunately, you fell in the same mistakes like others, you imposed that all can move with you to the next steps by only saying “repeat this steps” .. So I suggest to you more illustrations on why we repeat phase encoding gradiant by example of 4x3 matrix to distinguish between phase and frequency voxels differentiation steps.. Many thanks for your unique explanation method and for your patience..
Thank you for the feedback! These lectures are meant as a way to intuitively get a grasp of how we build an image and what phase and frequency encoding are while avoiding all the math which is not insignificant. So this is more of a working framework to get a good base understanding from, but not an exact explanation and I encourage you and everyone to think critically as to how this may not explain the full story as a technologist or radiologist would understand it knowing the depth and variations of the sequences we perform on a daily basis. The full, complete explanation will be covered in future lectures. Stay tuned!
I have watched several videos and read material from various sources on phase encoding. Your content has been the easiest to comprehend which otherwise is a challenging topic. The other sources either don't go into much detail or they go too much into the weeds. You have kept it simple with good animation and even the math has been simplified so it is easy to understand. I commend and appreciate your efforts, your knowledge and presentation skills. Quick question - do you plan on doing a presentation on Receive Bandwidth, sampling time, relation to frequency FOV, Nyquist theorm etc. Oh, and K space too :D
Thanks so much for the comment! I definitely have a whole set of lectures dedicated to k-space and how we truly build an image which involves getting into the math but I think I can present it in a visual heavy way still that will be intuitive. Not sure how deep we'll go into Nyquist theorem and such, always trying to find that balance between keeping it understandable without going too deep into the weeds but we will see! We very briefly touched upon the idea of Receive Bandwidth, I believe in the frequency encoding lecture but if you have more questions please send them my way! ua-cam.com/video/DYj1SLNppQM/v-deo.html
It's a shame that channel like this, get this many views, while "educational" videos in this subject get way more. Finally understood the situation fully before Medical Imaging Techniques retake exams. Keep up the good work doc! 😎
Thank you! Yes the whole algorithm thing has been frustrating, it seems like no matter how many subscribers I get or how much I advertise on social media, every video is just a slow creep from the moment it's released until it slowly makes its way up the ladder with views and likes. Maybe UA-cam isn't a fan of my jokes 🤷 But thanks for commenting and please do share with anyone you may think would find these helpful! Good luck with the exams!
Like, I could get through test just by reading the class's material and memorize them, however for the life of me I couldn't figure out how the process actually works, and it bothers me greatly. Thank to your video that studying is a lot more fun for me now, since everything begins to clear up now!
Thanks! Assuming we have e.g. a 100x200 pixel image in the xy-plane. Do we have to perform 100 individual measurements (with 100 different phase shifts and therefore 100 different applied gradients in the y-direction) in order to solve a 100x100 linear system of equations because we have 100 unknown amplitudes?
Love this question. The short answer is yes, this the classic teaching you will find. But this seems a bit... unrealistic doesn't it? Unfortunately, to understand what is really going on here, we need to dive deep into the math governing image reconstruction which we will do in future videos :)
This perplexing, I got to know how to calculate it when there is only a phsse shift, but how do we go calculating it when there is also a frequency shift at the time we measured the signal
The frequency localizing gradient is not a shift but an encoding of different frequencies along the x-axis. Since we create this gradient magnetic field, we know which frequencies should be where spatially along the frequency encoding axis, so we when we break our raw signal down into frequencies via the Fourier Transform, we get this information. Check out the dedicated frequency encoding lecture here for further info if you haven't already: ua-cam.com/video/DYj1SLNppQM/v-deo.html
@08:51 the addition and subtraction of By magnetic field is a bit confusing.Looking at the gradient on Y axis, By should be subtracted in the lower rows and added in the upper rows but you demonstrated the opposite.please explain thanks
Good eye and I debated on what is the best way to portray two gradients simultaneously on the image because it definitely can get confusing. I decided the most accurate would be to orient the addition and subtraction based on the axis the gradient is applied across. If you look at the X axis gradient, it is below the axis on the left (subtracted from B0) and above the axis on the right (added to B0). If you now turn your head 90 degrees counter-clockwise and look at the Y-axis, the portion of the gradient along the Y-axis in the bottom left voxel is above (added) and below in the top left voxel (subtracted). Hope this helps!
@@MRIPhysicsEXPLAINEDI also have the same query… not convinced by your answer 🤔🤔… u said we should turn around our heads in counterclockwise direction to see the y axis gradient… why not clockwise? In that case we will find +By in the top voxels and -By in the lower ones.. following the same rule as for the X axis…please explain.. thanks
@@rubyamir5245 Hello, sure you can turn your head the opposite direction on any coordinate system and the sides will be flipped, but mathematically you essentially get the same result albeit a change in signs. The important point is that we are creating a linearly changing magnetic field along the Y-direction of the slice, it does not really matter whether it is increasing going down to up or up to down.
Hello, check out the lectures on slice selection ua-cam.com/video/v8jW8K1y-KE/v-deo.html and frequency encoding ua-cam.com/video/DYj1SLNppQM/v-deo.html where we go in-depth on how to localize the signal in the Z and X dimensions in you haven't already. If you still have questions let me know!
@@MRIPhysicsEXPLAINED I'm confused about why we don't get the same problem at 7:55 in the Z-X plane. I can see that using frequency gradient in the X and Y axes causes the same frequencies to repeat and thus we can't resolve along (x+y=c). But why do we not get the same problem when using frequency gradient in both the Z and X axes? How are we still able to resolve voxels along those diagonals?
The beautiful thing about the slice select gradient technique is that we don't have to resolve the signal in the z direction, the gradient is turned on along the z-axis so that when we apply our RF pulse and tune it to the region of the body of interest along the z-axis, we know all the signal we receive is coming from that slice alone. We simply just then turn off the z-axis gradient, the protons in the slice fall back into precession at the Larmor frequency, and the problem then turns simply into localization along the x and y axes for which we do the frequency encoding and phase encoding techniques. Now in 3D MRI acquisition we do replace this slice select gradient technique with a true image encoding gradient along the z-axis, but that will require a little extra knowledge to understand which we'll cover in future lectures 🙂
Amazing videos, but I was a bit confused about the algebra portion. How did you know which voxel, in the column was CSF and which was soft tissue? How did you know that A was bigger than B? Couldn’t B have been bigger than A? In other words, how did you know that the Soft Tissue was on the bottom and the CSF was on the top? Why not the reverse?
These are great questions! Is this coming from after seeing the newest lecture? If not, check it out now as we talk about this very issue! ua-cam.com/video/ANUDUGg4F1c/v-deo.html
so why would you need k space if the signal can be decomposed into separate waves from each frequency/phase combination representing a point in cartesian space. the whole purpose of k space is that the signal at each point in cartesian space is a superposition of weighted waveforms of differring frequencies. if each point in space represents a specific frequency and phase, then there is no k space
My friend you are way ahead of the learning curve! The core lecture series is developed for the everyday radiologist and technician who wishes to get a feel for MRI physics without getting into the math which is multivariate calculus level. This expands on the concepts taught in most paid MRI physics courses, and as you have pointed out, what is the point of K-Space then? This is why K-space can be such a confusing topic in these courses, and in order to understand it and how we truly build a picture, we have to dive into the actual mathematics governing MRI image generation and we will do that in the future advanced series :)
Livesaver, i could not for the love of god figure out the phase encoding gradient like ever. After this i realised we went over phase changes back in middle school and with the help of ur vids it completed the picture for me. Will definitely recommend u to everyone studying this. ❤
Thank you so much! So happy to hear it helped demystify the whole phase encoding thing!
This content is just amazing. I have spent hours upon hours watching videos on other channels and reading on websites like MRI Questions and still felt unsettled and unable to integrate the various concepts. After watching this, I feel SO MUCH better and feel I now have a solid conceptual understanding (especially understanding how the Fourier transformations help in identifying unique spatial frequencies). I wish I watched this at the start as it would have saved me such an awful headache! Thank you so much!
Honestly comments like these are the greatest reward for making these videos, happy to hear it helped make things click. Credit to MRIQuestions.com and Dr. Elster for the underpinning theory here, I just expanded on it a little :) And while the real picture building process is a little more complicated, this ties in well with the typically taught base MRI physics concepts while avoiding the math. If you're really curious to how the real thing works, stay tuned! We will have a full series on it in the future including K-space.
This is one of the best materials of MRI physics I have encountered. Thank you so much for your time and effort in making this.
So glad it was helpful, have a lot of new material in the works so stay tuned!
Now I see the world with different eyes. Brilliant explanation
Thank you, happy to hear you found it helpful!
Life has been good to me. This explanation has finally gotten inside my head. Excellente job doctor!
So happy to hear it helped and thanks for supporting!
This was a lot to take in...I def need to watch this a couple more times for it to totally soak in. With that said this instructor is amazing
What is PHASE?
Imagine we are singing a song together and there are three different possible variations:
1) One of us is louder or quieter than the rest, that is a matter of Amplitude.
2) Amplitude of our voices are the same but one on us is singing faster or slower that the rest, that is a matter of Frequency.
3) Amplitude of our voices are the same (nobody is louder or quieter that the rest) and Frequency of singing are also the same (nobody is singing faster or slower the other people), but one of us is a little ahead or behind than the rest, that is a matter of PHASE (Time, but it is measured in angular degree, from zero to 360).
Love it!
This video is certainly underrated.
Hard to be the new kid on the UA-cam block. But liking, commenting, and sharing the videos will help spread the word, thanks for the support!
I'm taking the ARRT soon and these videos have helped me so much! I feel I finally understands the physics. Thank you for your lectures!
So glad to hear, good luck on your exam and let me know how I can help further!
Awesome, will be sharing this with my radiology trainee colleagues here in the UK :)
Thanks so much!
Thanks for sharing and supporting and cheers from across the pond!
Brilliant video describing how phase shift works, best in class.
Thanks for the support and watching, happy it helped!
thank you so much for your work. To be able to explain this well you had to understand it so well yourself and then trace back the process to each logical step! impressive!
You're very welcome and glad it was helpful!
Excellent and easily apprehendable way of presenting such a complex matter, Dr. TE! Thank you very much!
Happy to hear it helped, thanks for the support!
So, so, so good! Please have the T1 and T2 contrast videos asap 😊
Haha in the works, thanks for the support!
This is COOL! You are a great teacher!
Thank you!!
It was easiest explanation I have encountered till now ,Plz post more videos
More are in the works!
Before this moment 20:59 I was very interesting by your fantastic explanation and pretty animations, but suddenly you go through the nexts..
Unfortunately, you fell in the same mistakes like others, you imposed that all can move with you to the next steps by only saying “repeat this steps” ..
So I suggest to you more illustrations on why we repeat phase encoding gradiant by example of 4x3 matrix to distinguish between phase and frequency voxels differentiation steps..
Many thanks for your unique explanation method and for your patience..
Thank you for the feedback! These lectures are meant as a way to intuitively get a grasp of how we build an image and what phase and frequency encoding are while avoiding all the math which is not insignificant. So this is more of a working framework to get a good base understanding from, but not an exact explanation and I encourage you and everyone to think critically as to how this may not explain the full story as a technologist or radiologist would understand it knowing the depth and variations of the sequences we perform on a daily basis. The full, complete explanation will be covered in future lectures. Stay tuned!
I have watched several videos and read material from various sources on phase encoding. Your content has been the easiest to comprehend which otherwise is a challenging topic. The other sources either don't go into much detail or they go too much into the weeds. You have kept it simple with good animation and even the math has been simplified so it is easy to understand. I commend and appreciate your efforts, your knowledge and presentation skills. Quick question - do you plan on doing a presentation on Receive Bandwidth, sampling time, relation to frequency FOV, Nyquist theorm etc. Oh, and K space too :D
Thanks so much for the comment! I definitely have a whole set of lectures dedicated to k-space and how we truly build an image which involves getting into the math but I think I can present it in a visual heavy way still that will be intuitive. Not sure how deep we'll go into Nyquist theorem and such, always trying to find that balance between keeping it understandable without going too deep into the weeds but we will see! We very briefly touched upon the idea of Receive Bandwidth, I believe in the frequency encoding lecture but if you have more questions please send them my way! ua-cam.com/video/DYj1SLNppQM/v-deo.html
You are a godsend! Thank you so much.
My pleasure! If you're interested in more physics/radiology education, check out our new project radiofreedia.org (work in progress)!
It's a shame that channel like this, get this many views, while "educational" videos in this subject get way more. Finally understood the situation fully before Medical Imaging Techniques retake exams. Keep up the good work doc! 😎
Thank you! Yes the whole algorithm thing has been frustrating, it seems like no matter how many subscribers I get or how much I advertise on social media, every video is just a slow creep from the moment it's released until it slowly makes its way up the ladder with views and likes. Maybe UA-cam isn't a fan of my jokes 🤷 But thanks for commenting and please do share with anyone you may think would find these helpful! Good luck with the exams!
@@MRIPhysicsEXPLAINED Thank you very much, my whole year already watched your videos, so dont worry, you'll get there eventually 😁
@@mateuszskrzypczyk3563 Awesome thanks!
This is truly Amazing! Thank you so much!
Thanks for watching and commenting!
I have to watch these videos more than once.
Me too 😂
Great thanks
Excellent!
Thanks for watching and commenting!
Amazing lecture. Thanks so much!!!
Thanks for commenting and supporting the channel!
Like, I could get through test just by reading the class's material and memorize them, however for the life of me I couldn't figure out how the process actually works, and it bothers me greatly. Thank to your video that studying is a lot more fun for me now, since everything begins to clear up now!
So great to hear and happy I could help make this process a little more clear and hopefully entertaining! Check back tomorrow for the latest video!
Amazing!
Glad it helped!
Nice work!
Thanks for the support!
Thanks! Assuming we have e.g. a 100x200 pixel image in the xy-plane. Do we have to perform 100 individual measurements (with 100 different phase shifts and therefore 100 different applied gradients in the y-direction) in order to solve a 100x100 linear system of equations because we have 100 unknown amplitudes?
Love this question. The short answer is yes, this the classic teaching you will find. But this seems a bit... unrealistic doesn't it? Unfortunately, to understand what is really going on here, we need to dive deep into the math governing image reconstruction which we will do in future videos :)
Please make videos on k space and how everything fits in with respect to signal processing
In the works, thanks for watching!
Thanks for ur effort . Waiting for more 😅😅😅
Thanks for supporting! Check out the new video that just dropped today :) ua-cam.com/video/1ljzLjqgOGc/v-deo.html
This perplexing, I got to know how to calculate it when there is only a phsse shift, but how do we go calculating it when there is also a frequency shift at the time we measured the signal
The frequency localizing gradient is not a shift but an encoding of different frequencies along the x-axis. Since we create this gradient magnetic field, we know which frequencies should be where spatially along the frequency encoding axis, so we when we break our raw signal down into frequencies via the Fourier Transform, we get this information. Check out the dedicated frequency encoding lecture here for further info if you haven't already: ua-cam.com/video/DYj1SLNppQM/v-deo.html
@08:51 the addition and subtraction of By magnetic field is a bit confusing.Looking at the gradient on Y axis, By should be subtracted in the lower rows and added in the upper rows but you demonstrated the opposite.please explain thanks
Good eye and I debated on what is the best way to portray two gradients simultaneously on the image because it definitely can get confusing. I decided the most accurate would be to orient the addition and subtraction based on the axis the gradient is applied across. If you look at the X axis gradient, it is below the axis on the left (subtracted from B0) and above the axis on the right (added to B0). If you now turn your head 90 degrees counter-clockwise and look at the Y-axis, the portion of the gradient along the Y-axis in the bottom left voxel is above (added) and below in the top left voxel (subtracted). Hope this helps!
@@MRIPhysicsEXPLAINEDI also have the same query… not convinced by your answer 🤔🤔… u said we should turn around our heads in counterclockwise direction to see the y axis gradient… why not clockwise? In that case we will find +By in the top voxels and -By in the lower ones.. following the same rule as for the X axis…please explain.. thanks
@@rubyamir5245 Hello, sure you can turn your head the opposite direction on any coordinate system and the sides will be flipped, but mathematically you essentially get the same result albeit a change in signs. The important point is that we are creating a linearly changing magnetic field along the Y-direction of the slice, it does not really matter whether it is increasing going down to up or up to down.
Thank you very much pro
Thanks a lot. Extrmely good
Thanks for watching and commenting!
Great jobs
Thank you!!
sooo goodddd
Glad it helped, thanks for commenting and supporting the channel!
Question: how were we able to localize in the Z-X plane before when we were also using gradient magnetic fields in both these axes?
Hello, check out the lectures on slice selection ua-cam.com/video/v8jW8K1y-KE/v-deo.html and frequency encoding ua-cam.com/video/DYj1SLNppQM/v-deo.html where we go in-depth on how to localize the signal in the Z and X dimensions in you haven't already. If you still have questions let me know!
@@MRIPhysicsEXPLAINED I'm confused about why we don't get the same problem at 7:55 in the Z-X plane. I can see that using frequency gradient in the X and Y axes causes the same frequencies to repeat and thus we can't resolve along (x+y=c). But why do we not get the same problem when using frequency gradient in both the Z and X axes? How are we still able to resolve voxels along those diagonals?
The beautiful thing about the slice select gradient technique is that we don't have to resolve the signal in the z direction, the gradient is turned on along the z-axis so that when we apply our RF pulse and tune it to the region of the body of interest along the z-axis, we know all the signal we receive is coming from that slice alone. We simply just then turn off the z-axis gradient, the protons in the slice fall back into precession at the Larmor frequency, and the problem then turns simply into localization along the x and y axes for which we do the frequency encoding and phase encoding techniques. Now in 3D MRI acquisition we do replace this slice select gradient technique with a true image encoding gradient along the z-axis, but that will require a little extra knowledge to understand which we'll cover in future lectures 🙂
amazing, thanks! new sub here!
Thanks for the sub and supporting the channel!
Amazing videos, but I was a bit confused about the algebra portion.
How did you know which voxel, in the column was CSF and which was soft tissue?
How did you know that A was bigger than B? Couldn’t B have been bigger than A?
In other words, how did you know that the Soft Tissue was on the bottom and the CSF was on the top? Why not the reverse?
These are great questions! Is this coming from after seeing the newest lecture? If not, check it out now as we talk about this very issue! ua-cam.com/video/ANUDUGg4F1c/v-deo.html
@ no I haven’t gotten that far yet. I guess something to look forward to! Thanks
Thanks alot ......
Most welcome, hope you found it useful!
thank you
Welcome!
so why would you need k space if the signal can be decomposed into separate waves from each frequency/phase combination representing a point in cartesian space. the whole purpose of k space is that the signal at each point in cartesian space is a superposition of weighted waveforms of differring frequencies. if each point in space represents a specific frequency and phase, then there is no k space
My friend you are way ahead of the learning curve! The core lecture series is developed for the everyday radiologist and technician who wishes to get a feel for MRI physics without getting into the math which is multivariate calculus level. This expands on the concepts taught in most paid MRI physics courses, and as you have pointed out, what is the point of K-Space then? This is why K-space can be such a confusing topic in these courses, and in order to understand it and how we truly build a picture, we have to dive into the actual mathematics governing MRI image generation and we will do that in the future advanced series :)
Hello, am an MR physicist and these videos are superb, do you have a patreon or paypal?
EDIT: NVM just saw your description, enjoy the coffees
Wow thank you so much, it really means a lot!!
Pretty even matchup, but I think k-space takes home the prize money.
😂
Thank you y
Glad it helps and thanks for the comment!
😍😍😍😍😍
😁
10:52 0_0
Didn't know the former president was such a big MRI physics fan and I apologize to the nation for such bad jokes! 😂