How Bandwidth Affects Signal to Noise Ratio (SNR) in MRI | MRI Physics Course #12

[@kayk928]

Thank you so much you dont know how much this helps me as an MRI student. I rewatch your videos multiple times a day everyday! Please continue posting MRI contents!

[@radiologytutorials]

I'm so glad you find the videos helpful! Working on some more content right now 🙌

[@simonjohnson9865]

Thank you for your excellent work. Your teaching style is second to none.

[@radiologytutorials]

Cheers Simon. That means a lot. Thanks mate 🙂

[@kostiantyn6526]

Thanks!

[@radiologytutorials]

Thank you so much ☺️

[@sid17391]

Thanks a lot for making MRI miles more understandable than most other explanations.

[@lhyllianamarseilles]

The way you explain and illustrate these concepts is truly amazing. I can visualize my learnings because of your videos. Thank you, Sir Michael. I will be taking the ARMRIT soon, I hope the MRI Question bank will be ready in a few days.

[@radiologytutorials]

Thank you! I hope your studying has gone well 🙂

[@fazaltahir4216]

Words are not enough to appreciate your efforts, u made MRI physics so interesting n easy, big applause 👏 🎉❤

[@sid17391]

really hope this series and the mri q bank is all ready before the part 1 in September.

[@alaaalshammari5699]

God blessing you... عاشت ايديك يابطل

[@boomtown1020]

Excellent content and clear explanations, Thank you! Can you tell me when the MRI question bank will be available?

[@radiologytutorials]

Thank you @boomtown1020! Working on it at this very moment - hoping it will be available soon 🤞

[@christianseiler7132]

Hello Michael, you have a great way of explaining things! But I have a question regarding this video. I understand how a higher BW shortens the echo duration and that you need a higher sampling rate. Therefore you´re faster collecting all your data points you need. But wouldn´t it be possible to just use the same high sampling rate for a lower BW. Then you would need the same short time? Or asked in another way. Why not always using the highest sampling rate that is technically possible to be faster or at least have more data? Where is my mistake? Thanks in advance!

[@radiologytutorials]

Hi Christian. Excellent question! You’re one step ahead. Indeed what most sequences do is oversample in the frequency encoding direction. This helps to prevent aliasing occurring because we sample at a high enough rate to accurately localise signal that is beyond the nyquist limit for the FOV. So there is definitely benefit in using a higher sampling rate. There are limitations though. 1. Machines have a sampling rate limit and nearing the limit can make data acquisition less accurate. 2. Sampling at a higher rate does not speed up acquisition time - the echo still takes longer in lower bandwidths irrespective of sampling rate (which is also a good thing because we get better signal to noise). Hope this makes sense 👍🏼

[@christianseiler7132]

@@radiologytutorials Thanks a lot for the detailed answer! So you always have to sample the whole echo, if I understand it correctly. It is not possible just to sample the middle part of an echo with low BW to save some time?

[@radiologytutorials]

You can just sample the middle part (provided you have enough samples) but it wouldn’t save time, because you still have to wait until the TR and repeat the process for multiple phase encoding steps. Only processes that reduce TR/phase encoding steps/NEX will actually save time. Hope that makes sense 🙂

[@suman1650]

Are you planning to bring for CT also?

[@RobertWilliamsserving]

Wow this is sooo helpful thank you so much!!

[@radiologytutorials]

Glad it was helpful!

[@islamshawqi3823]

Sir , what's the difference we need to take in consideration with the bandwidth when working on open MRI machine ( weak magnets example: 0.3 T ) and when working on strong closed MRI machine ( 1.5 T ) What consideration we need for the Bandwidth parameters in order to get good SNR images , and thank you in advance sir.

[@fazaltahir4216]

Thanks a lot sir Michael 🎉

[@radiologytutorials]

Pleasure Fazal. Hope you’re well 🙂

[@fazaltahir4216]

Sir we are desperately waiting for new video of MRI physics,,,,

[@fazaltahir4216]

​@@radiologytutorialssir Michael we are waiting for the new video of MRI 😂, any updates plzz 🙏

[@radiologytutorials]

Hoping to film a new one tomorrow. I’ve been extremely busy the last 10 days. Just had no time to sit down and film 😣😣

[@fazaltahir4216]

​@@radiologytutorialsstay blessed sir Michael,,, your videos are very very helpful for us ❤🎉😊

[@sohailkohri7269]

Thankyou sir🎉🎉

[@radiologytutorials]

Cheers Sohail 🙏🏻

[@chazsison]

Sir how many videos to go before ct module?:)

[@RobertWilliamsserving]

Question: does the $14 review questions only apply to current video or does that purchase cover questions for all the sections for this series? Thank you!

[@radiologytutorials]

Hi 👋 Sorry for the delay! Missed this comment. The review questions cover topics covered in the entire MRI series.

[@RobertWilliamsserving]

@@radiologytutorials awesome! Thank you! Incredibly helpful I'll be getting those

[@krazykrish3uk531]

Hi, your videos are usefull.can you make video on pulse sequences like diifferent types of all pulse sequences in mri . Can you please tell about cardiac ,fuctional and diffusion tensor imaging in mri.

[@radiologytutorials]

Thank you. Will be making pulse sequence videos next week 😊

[@gudurisuresh7431]

Thank you doctor

[@radiologytutorials]

My pleasure 🙂🙏🏻

[@abdou.b3259]

Do you have books you suggest to understand quantum mechanics and calculus?

[@tomare6479]

Type in pdf syllabus… xyz. However, applied explanations are sometimes best.

[@tahirkohri4282]

Amazing ❤

[@SayraBano-l3l]

Best explain

[@radiologytutorials]

Thank you 🙏🏻

[@bodanamusicproduction876]

Amazing🎉

[@radiologytutorials]

Thank you! 🙌🏼

[@ayaankhoriblock5699]

Great❤❤

[@tysonbrown543]

I'm pretty sure 30khz bandwidth will give you 4.3ms - dont forget the nyquist theorem. Also the readout prephase gradient should be before the 180 to avoid adding the FID from the 180 😊

[@radiologytutorials]

Hi Tyson. I’m not sure I’m following your calculation. Are you referring to 4.3ms for the sampling time (in a 30khz bandwidth slice with 256 pixels)? A 30khz bandwidth slice has a nyquist limit of 15khz because the relative frequency changes are measured from the null point. These formulas take into account the nyquist limit. Perhaps I’ve made a mistake and am happy to hear you explanation and calculations.

[@tysonbrown543]

Ok I guess it depends on how you frame it. For us, if the operator requests 30kHz readout bandwidth, our acq time (s) = read points * (1 / (30000*2))

[@GajalaBano-k6v]

I like that❤❤

[@radiologytutorials]

Thank you ❤️

[@ayeshaishrat1861]

​@@radiologytutorialsplz put new cases also...u teach in very simple and most understandible way...

[@abi077]

Hi sir❤

[@mehboobalimehboobali5528]

Wow❤❤

[@radiologytutorials]

❤️

[@troyfrei2962]

Hold on, If the band with is smaller then the Image HiGHER freq will be CUT OFF. There for the RIGHT SIDE OF THE IMAGE MUST BE CUT OFF.

[@EstelleBuxton]

9344 Reynolds Mountain

[@AainaKohri]

❤❤

[@radiologytutorials]

❤️