MRI Physics

Basics

  • Protons have a positive charge and normally spin at baseline

    • When there is a moving electrical charge this creates a magnetic field (physics property you just have to know and accept)

  • When the proton is put into an external magnetic field the protons will do one of 2-3 things

    • Align themselves parallel to the magnetic field

    • Align themselves anti-parallel to the magnetic field

    • Note that being in parallel and anti-parallel require different amounts of energy/energy levels to be able to do this

      • The protons will preferentially align with the position that requires the lower energy

        • However a stronger magnetic field can allow the protons to go to the higher energy alignment

    • The following example is great:

      • People can walk around their house on their feet like normal, or by standing on their hands and walking upside down

        • Obviously walking on our feet would be easier and require less energy so this will be the preferred alignment

          • However if I give you an energy drink you may feel energized to walk on your hands for a period of time

            • The hand vs foot walking is the parallel vs anti parallel orientation

            • The energy drink is the strength of the magnetic field

  • It should also be noted that the alignment and pull of the protons in their respective directions is additive

    • If there are 10 protons processing upward (parallel) and 5 processing downward (anti-parallel) the net magnetic field will be upward

    • This concept applies for the X-Y-Z axes

  • SO…

    • When we put the protons in your body into an external magnetic field (MRI machine) the protons will align in the direction of the magnet (MRI machine)

      • The issue is that we cannot measure the strength of the magnetic force of just the protons in the body because they are in the same direction as the larger magnetic field made by the MRI machine which basically covers the small one formed by the body

        • So what can we do about this?

          • We can change the direction of the magnetic field so that it is transverse to the body and then record the difference

      • Big Takeaway is that we cannot measure the magnetization that occurs longitudinal to the magnetic field

    • Another great example:

      • A man is on a boat with a watering hose

        • If he points the hose into the ocean there is no telling how much water has been added to the ocean

        • However if we change the direction of the hose and point it to land then we can measure how much water is coming out of the hose

Procession

  • The protons will not just sit still within the magnetic field, they will move in a somewhat cylindrical/circular pattern which is called procession

  • Procession frequency is related to the strength of the magnetic field

    • Stronger magnetic field = faster procession = higher procession frequency

What happens when the patient is placed in the scanner

  • We send a radiofrequency (RF) pulse at the patient (radio wave)

    • This disrupts the peace of the protons

    • Note that not all RF pulses can be used, we need one that has the same frequency as the protons as this allows them to exchanged energy which then results in the disruption of the protons that we want

      • When the RF pulse and protons have the same frequency energy can be transferred between the two = termed resonance

        • Hence resonance in MRI

  • When the RF pulses gives energy to the protons it causes them to leave their low energy state (parallel/pointing upward) and become anti-parallel (point downward)

    • This results in a decrease in the magnetic field because now there are more protons pointing downward (anti-parallel) than before and this will causes a cancellation of the protons pointing upward (parallel)

Shimming/Shim Coils

  • Shimming makes the magnetic field (B0) as homogenous as possible

  • Reducing field inhomogeneities reduces artifacts with gradient echoes and chemical fat-saturation

    • Does not affect chemical shift artifact

  • In spin-echo it is the refocusing pulse that reduces field inhomogeneities not all this shit

  • How does it reduce field inhomogeneities

    • Passive shimming at time of installation (small strips of metal around the bore of the scanner)

    • Gradient offset shimming

    • Dedicated shim coils.

MRI AcquisitionTypes and Modifications

  • Now that a lot of the nitty gritty has been addressed we can talk about acquisition types and modifications

  • You should think of MRI as being acquired in one of a few ways (discussed here shortly) and then we can do several things (I call them upgrades here) to each of those acquisition types to get an additional perk/benefit

  • Acquisition types

    • Spin echo

    • Fast spin echo

    • Gradient echo

    • Echo planar

    • Steady state free procession

  • Upgrades

    • Weighting (T1, T2, etc.)

    • Fat saturation

MRI Acquisition

  • Spin Echo

    • Used a 180 degree rephasing pulse

      • Look at flip angle below for some more basic info on degrees and shit but for the 180 pulse i will explain below, but it it basically an UNO reverse card

      • Remember B0 is out baseline and everything is pretty much longitudinal signal

      • We will give a 90 deg pulse to put everything pretty much in transverse axis which then allows us to measure shit and get signal which can then become an image eventually

      • The spins will eventually start to decay and lose their synchrony however this will happen at slightly different rates due to micro differences in the magnetic field

        • Some are losing their spin faster than others basically, for example lets say we have 3 protons named A, B and C which are located between two posts labeled P1 and P2 as shown below with THE NORMAL PROGRESSION TO LOSE YOUR SPIN AND MOVE FROM P1 TO P2

        • P1 - A - B - C - P2, the normal progression

        • A has kept its spin well

        • B is somewhere in the middle

        • C is losing its spin quickly

      • We then hit the whole field with the uno reverse card we flip everything so that the slow are now the fast and the fast are now the slow and it looks like this

        • P1 - C - B - A - P2, so now A is the bad one and C is the best

      • At some point in time, the slow ones (C) will catch up with the fast ones (A) because we basically flipped the whole playing field to allow them to catch up (basically by allowing C to move less distance at the same slower rate and making A move farther with its faster rate…this is all to help understand, not actually moving distance its about spins)

        • The point where they protons all catch up with each other (eclipse if you will) means they are all in phase again which gives us a strong signal that can be detected by the scanner

      • Note that this 180 pulse is basically acting to get rid of the effect of the micro-differences in the magnetic field because it basically makes it an even playing field for all the protons because the ones that are moving fast get flipped and then move slow after the pulse and the ones moving slow are then fast after the pulse

        • Remember that the reason some move slightly faster or slower is because of micro-differences in the magnetic field

  • Gradient Echo

Upgrades

  • So what is weighting such as T1 and T2, etc.

    • When person is put in the machine the hydrogen protons in the patient will align with the main magnetic field —> this is called B0

      • Note that here at B0 the protons are all pretty uniform and are are synchronized in vertical and transverse spins

    • Then the machine will hit them with a RF pulse —> this tips the protons out of alignment with the main magnetic pulse they were originally aligned with

    • Once the pulse has occurred and the protons are knocked out of order, the protons will now do two things simultaneously

    • The longitudinal (vertical) spin will gradually return to the position it was originally (at B0) —> This is T1

      • If gets back to B0 quickly —> short T1 —> bright —> fat

      • If gets back to B0 slowly —> long T1 —> dark —> fluid

    • The other thing that happens is loss of synchronization of spins in the transverse plane = decay of transverse magnetization = T2

      • The spins are initially synchronized after the RF pulse and then will gradually lose synchronization

      • Harder to make the tissue lose synchronization = stays coherent longer = long T2 = bright = fluid

      • Easy to make the tissue lose synchronization = loses coherent quickly = short T2 = dark = solid tissue

    • w

  • What about TR & TE

    • TE = Echo time

      • This is the time from when you excite the tissue to when you measure the echo/signal

      • Remember you are going to hit the tissue with a RF pulse, wait, the echo will be there and then you will measure it

    • TR = Repetition time

      • You hit a single slice of tissue with the RF pulse and excite it —> acquire your signal —> then you excite it again

      • TR is the time between the two excitation pulses

    • So how do TE & TR fit into T1 & T2 weighting

      • The short answer is that T1 and T2 each have their own “range” of TR and TE which is what actually gives them their properties, as follows

      • T1 = short TR and short TE

        • A short TR means that we are hitting the tissue with repeated excitation pulses quickly

          • Tissues that require a longer time to get back to B0 in the vertical/longitudinal plane will not have been able to fully recover before the next pulse is sent out

            • Fluid takes a long time to recover = dark on T1

            • Fat has a quick recovery time = bright on T1

        • T1 emphasizes differences in longitudinal recovery basically

      • T2 = long TR and long TE

        • This is really emphasizing the long TE, which remember is the time from pulse to when the echo (signal) is measured

        • Tissues that keep their transverse magnetization (keep the spins synchronous longer) = long TE = bright = fluid

        • Tissues that lose/decay their transverse magnetization quickly (lose synchrony of spins fast) = short TE = dark = soft tissue

        • The long TR acts to minimize the effect of T1, whereas the long TE emphasizes the differences in transverse magnetization and therefore is basically what gives it the T2 characteristics

      • Ultimately, TR is what controls T1 weighting/characteristics and TE controls T2 weighting/characteristics

  • NOTE: THE TISSUES HAVE INHERENT PROPERTIES WHICH MAKE THEM THIS WAY, THE WEIGHTING IS JUST WHAT EXPOSES/SHOWS IT

  • Another important point, why do you need multiple pulses and not just one?

    • Each pulse basically gives you one data point, the more pulses, the more echos to measure the more data points at different locations in k-space which are all the used to make the image

  • Now lets talk about a flip angle

    • This is how far the RF pulse tips the magnetization away from the main magnetic field (B0)

    • At baseline the magnetization is in line at B0, which is predominantly longitudinal (minimal transverse magnetization so no signal present)

      • Not sure if this is by design or nature but it starts longitudinal

      • Once we flip the field transverse there is now a difference between B0 and where the magnetic field is, so now there is something that can be measured, and that is what gives us a signal and then subsequently an image

    • A 90 deg pulse tips completely into the transverse plane, perpendicular to B0 (because 90 deg angle)

    • A 30 deg angle will tip it 30 degrees and so on

    • Now again we have transverse (T2) and longitudinal (T1) magnetization

    • Bigger flip angle (90 deg)

      • More magnetization in the transverse plane = more signal immediately

      • Little longitudinal magnetization present —> can’t really use a short TR because tissue will not have enough time to recover from the large angle before the next pulse

      • More SAR = specific absorption rate = basically how much of the RF gets deposited into the patient as heat

      • Used in spin echo

    • Smaller flip angle (10-30 deg)

      • Less magnetization in transverse plane = less signal immediately

      • More longitudinal magnetization present = can use a short TR because will not take as long for tissue to recover before next pulse

      • Lower SAR

      • Used in GRE

    • Ernst angle - this is the optimal angle for a given tissue —> more math and formulas and shit than I am willing to memorize

  • Weights

Diffusion Weighted Imaging

  • B value

    • Higher b values/ADC values means

      • More signal loss = decreased signal to noise ration

      • Increases sensitivity to diffusion

      • Increased sensitivity to patient motion

Spin Echo MRI

  • Has a 180 degree rephasing pulse

  • Gradient refocusing refers to the bilobed frequency encoding gradient where a preparation part of the gradient dephases the spins, while the main part of the frequency encoding gradient rephases (refocuses) the dephased spins such that they are in phase in the middle of the readout. This refocusing is always done with the frequency encoding gradient

  • Less affected by field inhomogenities

Gradient Echo MRI

  • Do not have a 180 degree rephasing pulse

    • Therefore cannot have flow voids

  • More affected by field inhomogenities

References: