How Does an MRI Work and Differentiate Between Different Tissues?

[girts]

If this belongs to the high energy particle physics subforum please move it there. So since I have to take an MRI, I am curious to know how it works, I think I know the basics but the information about it out there is rather confusing.

So first off here is what I know. A large cylindrical magnet submerged in liquid helium to achieve it's superconducting state makes a uniform B field of high strength (1.5 or 3T) where the field lines are parallel to the axial direction of the cylinder (also the direction of human body from head to foot) , now here comes the difficult bit, as the human body is inside the MRI cylinder the strong B field aligns the proton spins parallel and some antiparallel to the field lines (as I assume that is their state of lowest energy being parallel to the external field.

Now there must be a set of different coils similar to the deflection coils in a crt tube located on the inner side of the MRI cylinder which when energized set up a B field that is perpendicular to the main external field, and I suppose yet another third set of coils similar to the second that make yet another b field which is now perpendicular to the second field? I assume these secondary field strengths are much lower than the primary axial B field?

I also understand from what I've read is that these secondary fields are there to change the alignment of the protons for a while and as those fields are switched off only the main field stays on and so the protons now want to get back to their lowest energy state and align themselves like before and as they do they emit a certain frequency wave which is then received by coils inside the MRI and the frequency is this signal can be differentiated and interpreted as either brighter or darker pixels on a screen which then illustrates different fluids and solids in a body which have different chemical properties.

Here is another thing I don't quite understand, how can the apparatus differentiate between protons in a belly fat and protons in say spine discs or bone matter? Because as far as I know all photons have the same properties like charge and spin and mass, so why would the protons in fat respond differently than the protons in bone matter if they are all in the same strength B field?Also could someone please explain the proton gyroscopic precession because in many videos the explanation involves saying that the proton physically spins about it's axis but I think that is a misconception isn't it?


this video also talks about a spinning proton which I believe is wrong?

So if the proton doesn't spin but it's spin is basically it's ,magnetic moment which is a property of it without it physically spinning then how does the precession takes place? does it simply wobbles around it's axis when influenced by external b fields without physically spinning around it's axis like a planet would do?

 

[sophiecentaur]

how can the apparatus differentiate between protons in a belly fat and protons in say spine discs or bone matter?

As I understand it, the decay time of the resonance is different for different elements and compounds so you can tell what was in the section that the MRI excited.

 

[muscaria]

I assume these secondary field strengths are much lower than the primary axial B field?

I think what is important is that they are pulse-like (short-lived) impulse fields which "tilt"/bring the spins out of equilibrium.

Here is another thing I don't quite understand, how can the apparatus differentiate between protons in a belly fat and protons in say spine discs or bone matter?

Again I am no specialist in this (so maybe someone else can add to this), but I believe there are a number of different relaxation-type times which can be used to image a variety of compounds. The spin-lattice relaxation time (T1) for instance has a large response for fat.

Because as far as I know all photons have the same properties like charge and spin and mass, so why would the protons in fat respond differently than the protons in bone matter if they are all in the same strength B field?

All of these spins are interacting generally with each other (spin-spin interactions) as well as with the "medium" (spin-lattice interactions) and the resulting signal depends on these interactions and the structure of the material etc.

 

[muscaria]

Also could someone please explain the proton gyroscopic precession because in many videos the explanation involves saying that the proton physically spins about it's axis but I think that is a misconception isn't it?

There is a classical analogy you can draw. It is an analogy which one shouldn't take too literally, but it still gives some insight. I remember seeing this video series a good while ago and enjoying it (it covers the analogy):

I can't remember if it mentions what it happening quantum mechanically, but we can get into that later if you want.

 

[ZapperZ]

There are many different procedures that can be used, but one of the most common, which has been vaguely described, is to induce a 90-degree flip of the magnetization. This will result in a spin-lattice relaxation time, whereby the bulk magnetization will decay back to its equilibrium value.

The reason why this is useful is that the decay rate for this process depends on the surrounding environment (the "lattice"). So different types of tissues will have different types of environment, so the decay rate will be different. This will allow the signal to be resolved for different types of tissues.

The other thing that is often done is that the external, static magnetic field may also have a spatial gradient, for example, along the length of the body. This results in a slight variation of the resonance frequency along different cross-section of the body. It allows for the operator to looks at various transverse slices of the body by picking up only the signal at a particular resonance frequency.

Zz.

 

[girts]

thanks folks for replying, well yes I had read about the relaxation phenomenon which is basically the time it takes for the proton to get back to the state of lowest energy and align with the large axial b field after it has been tilted by 90degrees with the help of the smaller high frequency B field coming from the secondary coils, correct?
But I read they speak about T1 and T2 relaxation times, well T1 is the time I just described but then what is T2? Is that another tilting yet in another direction to achieve some better radio frequency resolution of the picture?@ZapperZ , so the axial large magnet coil has a slightly varying strength along it axis so the relaxation times differ slightly from leg tissue to say head tissue and so one can slice up the body with a certain amount of slices much like higher bandwidth digital music track has more sound resolution than a lower one since it is sliced into more smaller parts?

Now I still don't get fully the "different tissue different proton relaxation times" fact, now we know that all protons are the same, so normally the whole body would come out as one dark or light blob on the screen but do you say that the nucleus in different fluids or tissue have different number of protons in the nucleus and the number of protons a given nucleus has impacts the time it takes for the protons to "relax" in that given nucleus with respect to another nucleus in other tissue given that both are screened in the same strength b field with the same frequency secondary field?
BTW, and I really want to understand this, so what's with the proton or electron for that matter spin? Is it just an old word that got stuck because at first when we discovered that these particles have charge and also magnetic moment we thought that they spin like little gyroscopes and that causes the magnetic moment and later found out that actually don't physically spin but simply have a magnetic moment as an inherent property much like charge?
Because if they don't physically spin that means they are simply like tiny bar magnets and they then align with the external b field (explains why metal can be magnetized in the presence of a b field) but this then begs the question once a secondary b field changes their position and is then switched off the protons process back to their previous state of lower energy but while doing so they behave like gyros which I assume is the process by which they emit characteristic RF by which they are then "seen" by the MRI?
Anyway what I'm asking is do they spin or not and if not then why would they process like a gyro upon such interaction because taking the analogy of a bar magnet it too can be aligned with an external b field and then its position changed but as it goes back to its previous lower energy state it does not rotate or process it simply flips back and that's it... this makes me confused.

 

[Dale]

Now there must be a set of different coils similar to the deflection coils in a crt tube located on the inner side of the MRI cylinder which when energized set up a B field that is perpendicular to the main external field, and I suppose yet another third set of coils similar to the second that make yet another b field which is now perpendicular to the second field?

The field produced by the gradient coils points in the same direction of the main magnetic field. Their job is not to change the direction of the field but to change its strength. Since the resonant (Larmor) frequency is proportional to the field strength, changing the field strength with the gradients changes the frequency.

 

[girts]

So its been some time and i have done the MRI scan so what i want to ask is this, while i was inside the torus shaped cylinder i felt stron pulsatin vibrations but the interesting part was that they were down well into the audible range , I'd say from few hundred hz to about 1khz judging from the pitch. So does the EM field they use for doing the proton precision change (sorry for vague techincal description) is in the low audible range of frequencies? Oh and another thing while i was inside in the moments that the pulses were done i felt a little bit to medium heating in my body and stomach which was also the part scanned and i think it wasnt a placebo effect could this be true?

 

[Dale]

So its been some time and i have done the MRI scan so what i want to ask is this, while i was inside the torus shaped cylinder i felt stron pulsatin vibrations but the interesting part was that they were down well into the audible range , I'd say from few hundred hz to about 1khz judging from the pitch. So does the EM field they use for doing the proton precision change (sorry for vague techincal description) is in the low audible range of frequencies?

Yes, that is correct. The gradient fields are switched on and off in the audio frequency range, and in fact they are powered by what essentially amounts to very expensive audio amplifiers, but these amplifiers are in the ~1 megawatt range! Everything possible is done to dampen the acoustic noise and vibrations, but when you are driving a 1 MW audio amplifier there is always going to be some acoustic and mechanical output.

Oh and another thing while i was inside in the moments that the pulses were done i felt a little bit to medium heating in my body and stomach which was also the part scanned and i think it wasnt a placebo effect could this be true?

Yes. The amount of RF energy that is deposited into the patient is very carefully monitored, and there are redundant safety systems, by law. However, the threshold for sensation is lower than the threshold for harm for many people, so feeling a little heating is common. It depends on the patient and the specific exam.

 

[girts]

Hmm just as I thought. Well then a question arises why use the very low frequency spectrum of the EM range for this proton precession thing? wouldn't they respond better by using higher frequencies and secondly using higher frequencies in the EM range would greatly simplify the apparatus as I believe in order to get a significant output power for an EM field in the audio spectrum you need a very powerful amplifier and a large coil as for the same amount of power using higher frequencies one could get away with a less powerful amplifier and smaller coil right?
The only thing I can think of is they use such low frequencies in order to minimize tissue heating?
oh another thing , so basically the main magnet is a static B field magnet made by generating a large current in a coil which is then kept at superconducting state but the field itself is static like from a permanent magnet and has a gradient along it's axial length, this main field simply aligns the protons in our body much like a magnet aligns electron spins in a metal, and then this pulsed low frequency EM AC type field disrupts the spins of the protons and makes them tilt as they do so the AC field is turned off and the protons again tend to align themselves back with the main static field but while doing so they emit some energy which they got from the secondary AC field disrupting them and this energy is radiated as EM field aka photons and received by the coils in the apparatus which then based on the strength of the field either paints a brighter or darker spot so forming an image which if done with high resolution can accurately show the structures of our inner bodies like bone matter and tissue and etc.
I wonder either the EM given off by our protons is kind of strong once inside such changing B field or the receiver coils are very sensitive because it is hard for me to imagine the blob that is our body to give off any significant radio frequency EM radiations simply from being subjected to a changing B field.
PS. I guess they can't minimize the noise much because it is both in the audio spectrum and essentially a sort of air core transformer since the secondary winding is meant to be our bodies right? even good large sized normal transformers made as tight as possible for efficiency emit quite a lot of mechanical noise at 50/60hz.

 

[Dale]

why use the very low frequency spectrum of the EM range for this proton precession thing? wouldn't they respond better by using higher frequencies

There are three types of magnetic fields in a MRI scanner, distinguished by their frequency range. The main magnetic field is static (DC), the gradient fields are in the audio frequency range, and the RF field is in the radio frequency range. The RF fields are used to excite the MR signal, the gradient fields do spatial encoding, and the static field sets the center resonant frequency. The strength of the RF and gradient fields are limited by adverse biological effects, so to mitigate those effects they have to go slower and that puts the gradient fields in the audio frequency range.

The only thing I can think of is they use such low frequencies in order to minimize tissue heating?

That limits the power of the RF field. The gradients are limited by peripheral nerve stimulation and cardiac muscle stimulation.

then this pulsed low frequency EM AC type field disrupts the spins of the protons and makes them tilt

No, that is the RF field.

I wonder either the EM given off by our protons is kind of strong once inside such changing B field or the receiver coils are very sensitive

They are very sensitive. The signal is microvolts, even with well designed receive coils located centimeters away.

 

[girts]

I wonder why are they using two additional signals one being the RF range and the other the low frequency one besides the main static DC field?
Is that in order to get a 3D model of the body? Say the main field being static can't give any info it just aligns the protons in order so that the two other fields can do any meaningful work on them with a meaningful signal coming back?

So the RF field precesses the proton spins in one direction while the other low frequency one then disturbs them in yet another direction and so working in unison these fields then can make the protons to give back EM signals from various positions in their precession unlike if they would use only one say the RF field the precession of he protons would only be in one axis?
I also wonder why are the two fields in such different frequency ranges? why aren't they both RF, is that due to some physical limitation with respect to the imaging itself or is that to limit the tissue heating and what you called "peripheral nerve stimulation" and "cardiac muscle stimulation" what exactly did you mean by these?
did you mean that if they used a stronger field or a higher frequency it would cause damage/disruptions in the heart's pace signal or something else?by the way how do you write the word precession?

thanks.

 

[Dale]

I wonder why are they using two additional signals one being the RF range and the other the low frequency one besides the main static DC field?

They serve different purposes. The RF field excites or inverts the spins, and the gradients do spatial encoding. The RF field has little or no spatial information, and the gradients neither excite nor invert the spins, and both are needed for making an image.

So the RF field precesses the proton spins in one direction

Yes, although the motion is technically a nutation. The main magnetic field makes the spins precess in the RF range. The RF field makes the spins nutate. The gradients spatially encode by making the precession rate depend on location.

while the other low frequency one then disturbs them in yet another direction

No, the gradient fields point in the same direction as the main magnetic field. They do not change the direction, they change the precession frequency. But the key is that they do so in a spatially dependent way. The main field and the RF field are both as spatially uniform as possible. The gradient fields are deliberately spatially nonuniform. So they map position to frequency, in other words they Fourier encode the spatial information.

working in unison these fields then can make the protons to give back EM signals from various positions in their precession unlike if they would use only one say the RF field the precession of he protons would only be in one axis?

The precession is only about the z axis (the direction of the main magnetic field). The RF, by itself, can cause nutation about either the x or the y axis. Again, that is not the purpose of the gradients. They do not change the direction of the spins, they do the spatial encoding by mapping position to frequency, a Fourier transform.

is that to limit the tissue heating and what you called "peripheral nerve stimulation" and "cardiac muscle stimulation" what exactly did you mean by these?

Switching the gradient fields induces eddy currents in the body. These eddy currents can stimulate electrically sensitive tissues. If it stimulates peripheral nerves then it can cause pain or involuntary muscle contractions. If it stimulates the cardiac tissue it can cause cardiac arrest. Patients prefer to avoid pain and spasms and they tend to not pay their medical bills on time when you kill them by cardiac arrest.

 

[girts]

nice touch of black humor there at the end, :D I liked it.

Ok so what I gather is this, the main DC field aligns the spins all into single direction in order for them to be useful at all, then the RF field is strong enough that when switched on it "distracts" (if I can say so) the vertical or along the z axis proton spins to some degree (90, more or less? with respect to the z axis?)
now as these protons are now in a precession to get back to align themselves with the main DC field a "scanning type" low frequency field is switched on and at the places where it imparts it's energy it kind of makes the protons to precess back faster than they would on their own since as you said the low frequency field is also along the z axis so what it essentially can do is make the protons align back faster but by aligning them back faster they emit a stronger signal (or higher frequency one?)correct?
so since the main field also has a gradient along the z axis, the low frequency field can then simply be dragged along the body inside the cylinder and at every horizontal "slice" along the z axis it can create an image but it is possible to stack the horizontal image slices up to make a correct representation of the body because each slice has a slightly different DC B field strength so the low frequency field gets slightly different responses at each slice from the spin alignment, is that how this works?
then the image is digitally processed and an image is formed?

and the reason why a black/white image can be formed similarly to a x-ray is because different tissues and even different parts of the same tissue type (like bone) have slightly different rates at which the protons align back if excited so they send a different signal (again is the signal of different frequency or strength or both?)thanks.

 

[Dale]

the main DC field aligns the spins all into single direction in order for them to be useful at all, then the RF field is strong enough that when switched on it "distracts" (if I can say so) the vertical or along the z axis proton spins to some degree (90, more or less? with respect to the z axis?)
now as these protons are now in a precession to get back to align themselves with the main DC field a "scanning type" low frequency field is switched on

Close enough. The motion is not a simple precession. There is precession, transverse decay, and longitudinal regrowth. That is necessary because simple precession could never realign with the main field.

at the places where it imparts it's energy it kind of makes the protons to precess back faster than they would on their own

It can make the precession go faster or slower, but it does not change the intrinsic decay or regeowth rates.

but by aligning them back faster they emit a stronger signal (or higher frequency one?)correct?

They do not align back faster. The gradient fields only cause them to precess faster or slower. It does not change the strength of the signal, only its frequency.

so since the main field also has a gradient along the z axis, the low frequency field can then simply be dragged along the body inside the cylinder and at every horizontal "slice" along the z axis it can create an image but it is possible to stack the horizontal image slices up to make a correct representation of the body because each slice has a slightly different DC B field strength so the low frequency field gets slightly different responses at each slice from the spin alignment, is that how this works?

This is way off. With the gradients the received signal is the Fourier transform of the image. Nothing is moved or dragged, it is just a physical mapping from position to frequency followed by a Fourier transform to obtain the image.

and the reason why a black/white image can be formed similarly to a x-ray is because different tissues and even different parts of the same tissue type (like bone) have slightly different rates at which the protons align back if excited so they send a different signal (again is the signal of different frequency or strength or both?

The signal strength determines the brightness of the pixel, the signal frequency determines its position.