Transcription
Hello everybody, and welcome back. We've made it to the final talk in this MRA module. Now, up until now, we've been looking at signal changes within blood vessels without the need for the administration of an exogenous contrast agent. Now, we're going to shift our attention to contrast-enhanced Mr angiography. We're going to see how giving contrast will allow us to generate more signal within blood vessels, and we're going to look at why we get more signal, as well as the type of sequence we need in order to capture that signal.
Now, if you've worked in MR, you'll notice that all of the contrast agents are generally gadolinium-based. Now, gadolinium has a unique property where it drastically shortens T1 times of tissues. Gadolinium generally stays in the extracellular space within blood vessels. It rarely crosses membranes. It's a large molecule, and it, in itself, is toxic. It needs to be bound to something known as a ligand in order to be able to be safely administered to our patients.
Now, why does gadolinium drastically shorten T1 times? Well, it's a unique atom in the way that its electrons are configured. You can see that six electron shells are filled within the gadolinium atom, and we don't just fill shells sequentially from the first shell all the way out to the sixth shell. We know that there's a rule called Hund's Rule, and it dictates the order in which we fill electron orbitals. Now, the way in which we fill electron orbitals means that there are certain orbitals within the fifth and sixth shell that are filled with electron pairs prior to the 4f orbital orbital within the fourth shell being filled with electrons. Now, the 4f orbital in gadolinium, because of the number of electrons it has, will have seven unpaired electrons – the maximum number of unpaired electrons that are able to be within a shell. Those seven unpaired electrons have a non-zero spin value. Electrons, in themselves, have a half-spin value.
Now, normally, when we fill an orbital, we fill the orbital with electron pairs – a spin-up and a spin-down electron. Those spins will cancel each other out, much, much like the protons that we've been talking about. These seven unpaired electrons don't have that canceling-out electron. They each have a non-zero spin value, and as a result, when gadolinium is placed into an external magnetic field, we get the same phenomenon happening. We get the production of a magnetic moment. We've got a non-zero spin, an external magnetic field, and the generation of a magnetic moment.
Now, because there are seven unpaired electrons and we've got a high g-magnetic ratio for these electrons, the gadolinium, in itself, will be paramagnetic. It will cause a change in the magnetic field. Essentially, what's happening is gadolinium will increase the amount of spin-lattice interactions, will cause hydrogen protons that have nuclear magnetic resonance to start regaining their longitudinal magnetization much more rapidly. Importantly, as well, gadolinium at high concentrations is also going to reduce the T2 time. It's going to cause rapid dephasing, but predominantly, the mechanism for generating more signal in contrast-enhanced MR angiography is that gadolinium rapidly reduces the T1 times.
So, let's have a look at an example here where we've got blood and soft tissue. What we want to do is we want to create contrast between blood that will have our contrast agent and soft tissue which won't have our contrast agent. Remember, the contrast agent isn't getting into the cells. So, if we give a contrast agent, what's going to happen is the contrast agent will only be within the blood vessels. Ideally, as we give that contrast agent, it's got gadolinium in it. Gadolinium is going to cause an increase in spin-lattice interactions, rapidly shortening the T1. We can see how blood with high concentrations of gadolinium is going to have this rapidly reduced T1 time. The T1 constants here have been separated. Remember, the T1 constant is a time taken to regain 63% of the longitudinal magnetization.
Now, if we were to have a sequence that actually showed the differences in these T1 times, what type of sequence would we need? If we had a very long TR time, both soft tissue and blood will fully regain their longitudinal magnetization, and we won't get separation based on these differences in T1 values. So, we need a short TR time. A short TR time is going to prevent soft tissue from regaining much longitudinal magnetization, whilst blood with gadolinium is going to have a lot of longitudinal magnetization. And remember, the degree of longitudinal magnetization, or the magnitude of longitudinal magnetization, will determine how much transverse magnetization we get at the next RF pulse. So, we need a sequence that has a short TR.
But remember, gadolinium as well also reduces the T2 times, so we want to have a short TE. We want to prevent it losing that signal after we flip the net magnetization vector into the transverse plane. Now, the ideal sequence for this is a gradient recalled echo sequence. Now, remember, gradient echo sequences have the advantage of being able to rapidly sample tissue with short TE and short TRs and get a very heavily T1-weighted image. And this sequence, specifically, we're using a spoiler gradient here. So, what we've essentially created here is an incoherent gradient echo image. We don't have that stimulated echo providing signal here, and we've got a T1-weighted image – a short TR with a short TE.
Now, we haven't touched specifically on 3D imaging, comparing that to taking multiple slices and stitching them together, but it turns out when we're doing contrast-enhanced MR angiography, we use a 3D technique. We sample a large volume of tissue, and we use slightly different slice selection gradients in order to separate where that signal is coming from. And that's something we'll perhaps touch on in a later talk. But doing a 3D image allows us to capture a large area of tissue over a short period of time. Speed here is of the essence. The gadolinium is going to run through these vessels quickly and only going to stay in the arterial system for a short period of time. It's then going to go into the venous system, and we're going to get veins overlying our image here. We're catching it in a phase before the gadolinium has reached the venous system.
So, if we're trying to detect the signal at a specific period of time, we need to be able to time when we actually acquire the signal from when we give the bolus to when we actually take the image. If we wait too long, we'll get that venous overlay. If we do it too quickly, we're not going to get the signal within the system that we're looking at. Now, there are multiple different mechanisms for timing this. The first is we can give a test bolus – a small amount of contrast – and see what that does to the image. How long after our test bolus do we start getting signal within the area of interest? Then we can give our full bolus and do the same timing, acquire the image when we have our signal coming through. Perhaps more frequently, what we can do is use what's called triggering. We can sample a specific volume where we know we want signal, and the machine can rapidly keep sampling this voxel here. And when we start to get signal coming in that voxel, a rapid increase in signal in that voxel, we trigger the sequence and then acquire it.
And the last mechanism we can use is what's known as time-resolved imaging, where we get a real-time look at the signal changes within blood vessels here. Now, this is slightly more complicated because we're not triggering when we want to acquire a specific still image here. We want to see how the image changes over time – the temporal changes. Now, the way we can do that is by manipulating how we sample k-space. We can acquire, without contrast, the region of interest. We can acquire that image. Now, the sequence that we're using, the gradient recalled echo sequence, incoherent sequence, we're using here uses a very short TR. It doesn't allow this soft tissue to regain much longitudinal magnetization prior to the next RF pulse, and as a result, we get partial saturation. We get very reduced signal from the soft tissue, and we can accentuate that by using a larger flip angle. The larger flip angle we use with this RF pulse, the less residual longitudinal magnetization we have, and we really get a suppression of the soft tissue signal.
Now, as we apply contrast, we are going to get high signal within the arterial system here. Why do we have high signal? Because with gadolinium, we get rapid recovery of longitudinal magnetization despite having short TRs. We still are recovering that longitudinal magnetization, and even though we've got a large flip angle, and actually because we have a large flip angle, we're getting a lot of signal at this short TE here. Now, if we were to sample k-space rapidly over and over again, we would have a lot of phase encoding steps required for each image that we take, and that takes time. Each additional phase encoding step here, we are adding time to our pulse sequence. So, what we can do is lay over that background image that we've taken prior to having contrast and only sample the center of k-space. If, if you remember when we looked at k-space, the center of k-space is what provides us with the most contrast information. It gives us the highest signal differences between different voxels within our image. We can rapidly sample the center of k-space using much fewer phase encoding steps. That allows us to get a real-time look of how signal is changing within the image, and that contrast is going to allow us to quite accurately do that.
Now, obviously, the more we try and get temporal resolution, the more rapidly we try and sample the center of k-space, we are going to have the side effect of losing some spatial resolution, and it's a trade-off there. The more we sample the entirety of k-space, the better spatial resolution we get, but we get a loss of temporal resolution.
Now, contrast-enhanced MRA has multiple different advantages. Because we're using a 3D technique here, we get a good signal-to-noise ratio. And as I've said, we get good temporal resolution. When we're looking at something like time-of-flight, where we're looking at intrinsic properties of the blood flowing into and out of the slice, we're not getting a real-time look of blood flow here. We also have good spatial resolution because of this 3D acquisition, and because it's a rapid sequence here, we're sampling so quickly as well that we get very little motion artifact here. We're not stitching multiple different slices together like we did in time-of-flight, and these images here are also less susceptible to flow artifacts. Remember, in time-of-flight, when spins travel through a slice or when there's turbulence, we're going to get either saturation or loss of signal. And lastly, in this technique, especially with peripheries, we can image quite a long section of blood vessels without risking saturation with our time-of-flight. As we increase the number of RF pulses that the blood is getting as we travel down that vessel, we might get some loss of signal due to partial saturation here. We don't have those effects because it's the gadolinium that's shortening the T1, and that's what's providing signal. It doesn't matter where the blood is along the vessel itself.
Now, obviously, administering contrast or doing a contrast-enhanced MRA technique comes with its own risks. Patients with chronic renal disease or chronic renal insufficiency are at risk of developing nephrogenic systemic fibrosis, and that's something that has been seen in patients who are unable to clear the gadolinium from their system. We also have incidences of acute reactions to gadolinium, which are mostly minor, but there are acute allergic reactions that can happen to gadolinium contrast, either the gadolinium itself or the ligand that it's bound to. And with the administration of gadolinium, we need to make sure that our timing is correct from when we give the bolus to when we actually acquire the image. If that timing is off, we're going to get contrast enhancement in the vessels that we're not interested in, either in the veins if we're looking at the arteries, or we're going to get contrast or signal coming from the ureters and the bladder that's going to interfere with our ability to accurately create a diagnostic image.
So, hopefully, you can see that the administration of gadolinium, and whatever form it is, is going to shorten the T1 times where there is gadolinium, generally within blood vessels. You may see in some neuroimaging that we get breakdown of the blood-brain barrier, and gadolinium can seep into the extravascular space. It doesn't become intracellular; it's still interstitial, but it's outside of the vessels. And where gadolinium is in high concentrations, it's going to be brighter signal. So, long as our timing is correct, gadolinium is going to shorten T1 times, improve longitudinal recovery, giving a bigger transverse vector at the RF pulse, and if we sample that transverse vector quickly before we lose T2 signal, before we lose transverse magnetization, we're going to get brightness where gadolinium is. Remember, it's not gadolinium itself that's providing the signal; it's still the hydrogen atoms or the hydrogen spins that are providing the signal, but they've gained longitudinal magnetization quickly because of those spin-lattice interactions.
So, that rounds off our MRA section. It's a large section, there's a lot to cover, and each section of MR angiography has slight differences, slight subtleties, and nuances within it. And it's important when you're going into an exam to have clearly in your mind how the different techniques differ from one another. Is a signal going to be dark within blood vessels, or is it going to be bright, and why do we have those signal changes within blood vessels? All of these types of questions I've included in a question bank link below. So, go and check that out if you're studying for a Radiology exam. Otherwise, I'm going to see you in the next talk. We're going to move away from MR angiography finally and move on to MRS, MR spectroscopy, and how we can see or assess the concentrations of different metabolites separate from water and fat, which we predominantly been looking at up until now. So, until then, I'll see you all there. Goodbye, everybody.