Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Medical How MRI scanners work

  1. Feb 25, 2010 #1

    I have been trying to figure out how MRI scanners work but am getting a bit confused and wondered if someone could put me straight on this ?

    I am just interested as they strike me as being amazing machines (possibly one of human kinds greatest inventions?) and the way they work is truly amazing (but pretty difficult to understand)
    I am not interested in the maths of it, just a general lay persons understanding of the basics
    BTW - If I ever manage to get some kind of understanding I may try to create a layman's explanation web page as I am sure a lot of people would be very interested in this

    I have watched these videos many times over (which are very good) - http://www.magritek.com/videos.html#01 [Broken]

    What I am stuck on at the moment is how it scans the patient (i.e. traverses K space ?)
    I keep thinking I am close to getting it then realise there is something I don't understand

    looking at the pulse sequence example on Wikipedia - http://en.wikipedia.org/wiki/Mri#Example_of_a_pulse_sequence
    This is what I think is going on:

    SS - puts a gradient on the patient (head to foot) and by careful selection of radio frequency only a thin slice through the patient at a specific point receives the 90 degree tilt (i.e. a slice is selected)
    So now any signal received will only be from this slice

    PE - puts a gradient across this slice for a set period of time to causes a phase shift across it
    (this seems to be suggested that its selecting a line but I dont understand how?)

    SS - Now a gradient is put across the slice in the other plane and again through careful selection of radio frequency only a thin line of the slice is given a 180 degree flip (to create a spin echo)
    So now any signal received is only going to be from this line through the patient

    FE - a gradient is then put across so giving a range of frequencies across this line relative to its location (readout gradient)
    the resulting signal is recorded forming a single line in K space?

    What I don't understand here is what is the phase needed for ?
    as the signals location seems to be given by its frequence so it is not needed
    Last edited by a moderator: May 4, 2017
  2. jcsd
  3. Mar 1, 2010 #2
    bumping after move
  4. Mar 2, 2010 #3


    User Avatar
    Science Advisor
    2017 Award

    The frequency can tell you the location of a signal only along one axis (e.g. the x-axis). You need the phase shift to tell you the location of the signal along the other axis (e.g. the y-axis).
  5. Mar 2, 2010 #4
    Thanks for the reply

    This is what I thought originally but then the timing diagram I was looking at on Wikipedia seemed to suggest that 2 slice selections were done so then I couldn't see why the phase would be needed (as you re only getting a signal back from a thin strip through the patient) ??

    I keep thinking I am getting close to getting it then read something which confuses me again ;-)

    I have just got myself a copy of "MRI in Practice" so I am hoping this will give me a better understanding (although from the look of the book it may just confuse me more ;-)
  6. Mar 2, 2010 #5


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

  7. Mar 2, 2010 #6


    User Avatar
    Science Advisor
    2017 Award

    The second "slice selection" step is acting to cause the spin echo on the same slice as was selected in the first pulse.
  8. Mar 3, 2010 #7
    Yes - that explains it - it now suddenly all makes sense
    thanks :-)

    It's one of those things which seems obvious now you tell me but I would never have figured it out
    I should have realised that the spin echo was involved in there - Doh!

    btw - I am sticking my head in a 7 tesla MRI in a few weeks (volunteering to help in a study) so this is what has got me so interested in them
    I am hoping to have a basic understanding of what is going on by then ;-)
    Last edited: Mar 3, 2010
  9. Mar 6, 2010 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    With a 7T MRI, don't be surprised to feel a little funny while you're in there. Nothing that's going to hurt you or last once the machine is turned off, but could worry or startle you if you don't expect it. We have a research group here that uses MRIs for their studies, and they've done their share of time in the machine, or near it assisting patients or subjects, and have told me there is a noticeable sensation when your head is being scanned.
  10. Mar 7, 2010 #9
    Thanks - yes, that would have been worrying as I have expected to feel nothing

    They are extraordinary machines.
    I had always assumed that something was physically moving round inside the MRI machine (something like C.T. scanners) but the idea that there are no moving parts but it scans just by varying the magnetic field and mathematics is amazing

    Then the way the resulting data (K space) is made up of each point describing a simple sine wave and the result of all these waves added together makes a complex image is staggering to me

    The way it "traverses k space" to collect the date is something I am still trying to understand.........

    Most machines you can get a very simple layman's understanding of how it works in 5 mins but with m.r.i. just this level of understanding is proving a challenge !
  11. Mar 18, 2010 #10
    Just to let you know I had my scan today and survived the ordeal ;-)

    I was scanned on both a 3.1 and a 7 tesla scanner
    I didn't feel anything (just a very slight flashing in my vision in time with some of the sequences on the 7 tesla) although someone else who has recently been reported feeling like "being pushed through a brick wall"

    I was surprised to find that the images were all ready to be viewed as soon as I got out of the scanners - I had assumed there would be a lot of time spent processing the data etc.?
    amazing images - I will be getting a copy on DVD soon so may post an example when it arrives

    BTW - Thanks for the help here
    I think I understand how they work just about enough now to satisfy my curiosity ;-)
  12. Apr 4, 2010 #11
    As promised, here are some of my scans:


    Hopefully any radiologists on the forum won't notice anything serious here ;-)

    btw - these are the images from the 3 tesla scanner - the 7 tesla ones were a bit disappointing (I suspect because they were just doing scans specific to their study?)
    This is from the 7 tesla scanner - http://alanesq.com/mri/7t.jpg
    Last edited: Apr 4, 2010
  13. Apr 5, 2010 #12


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Very cool pix :cool:!
  14. Apr 8, 2010 #13
    I have been seeing what I can do with the Dicom files
    e.g. producing 3d images of my brain, flashy animations etc.
    I have had some very limited success using 3D-Slicer
    http://www.alanesq.com/mri/head1.jpg and http://www.alanesq.com/mri/brain1.jpg

    but I find the program a real struggle (in fact since producing these I have not managed to do it again, I just get an error message ???)

    Just wondering if anyone here can recommend some easy to use software or has any suggestions of things I can try out etc. to keep me occupied ;-)

    BTW- the error I get is "ERROR: In ..\..\..\Slicer3\Modules\CommandLineModule\vtkCommandLineModuleLogic.cxx, line 1557
    vtkCommandLineModuleLogic (1E9FE588): Model Maker terminated with an unknown exception."
    Last edited by a moderator: Apr 25, 2017
  15. May 4, 2011 #14
    There does seem to be a profound difference between the 3 tesla MRI scanner and the 7 tesla MRI scanner. As far as understanding how the MRI works, I still am quite confused and impressed by its complexity. I need a MRI for dummies book to understand it completely.
  16. May 4, 2011 #15
    Don't be put off, it is very complex but truly amazing because of this, so worth spending some time trying to get your head round it all.

    I take my hat off to the people who design them etc......amazing stuff
    just trying to get a very basic understanding of them was stretching my mind to it's limit.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook