MRI Methods

MRI collects images but not all images are the same. The image types are called sequences because of the sequence of events that have to be programmed into the MRI to collect the data. Clinical exams include a T1 weighted sequence for instance.

To understand the range of images, you need some terms.

Spin just represents the fact that all nuclei are spinning. They spin at the frequency that is linearly dependent on the local magnetic field

Gradient is the gradient of magnetic field across the MRI. At “rest” the MRI has a uniform high magnetic field across the bore or tube. Lining the bore are room temperature electromagnets. Amplifiers put current into these electromagnets in order to create a magnetic field gradient across the MRI. These electromagnets are called gradients. The amplifiers powering them are gradient amplifiers. The gradient of magnetic field created is called, well.. a gradient.

     The gradient is key to the image. Every nucleus spins at a frequency that is specific to a magnetic field. If you put a gradient across the bore of the MRI, then as you cross along that gradient the spins are changing. After collecting the data, the computer does a fourier transform to figure out all the frequencies in the data. Since we know the field at any point in the MRI in space, and we know the frequency of the spin, we can place the intensity of that spin into the correct location in space. Gradients have many uses, but localization is a big one.

RF coil is the hardware (an FM tuned antenna) that is placed over the region of interest to collect signal (Head coil, body coil, wrist coil etc). This antenna transmits a local burst of energy which is what aligns the nuclei. This antenna detects the signal coming back. The signal is an induced current in the coil caused by the spinning nuclei (tiny magnets).

T1,T2,T2*. These are time constants (hence “T”). The signal coming back dies away quickly. The signal is lost through two processes. The signal in an MRI is based on the concentration of the nucleus (most MRI is based on water protons), and the time constants listed here. You will hear talking about T1 or T2 weighting, or T1 or T2 quantification. For technical info, see

The basics of MRI http://www.cis.rit.edu/htbooks/mri/

Wikipedia or the hundreds of MRI training sites.

Sequences. The gradients, RF coil transmission pulse and receive or acquisition electronics are all controlled by computer programs. These programs are called sequences as are the image method that they program.

Echo. In an MRI, signal first dies away due to the time constant T2 but they are still there in an orientation that can be detected. There is a trick, called an echo, which recovers the signal for a moment to allow it to be detected. You can create an echo using the RF coil (spin echo) or using the gradients in the MRI (Gradient echo).

So, here are some image types you can look up

Gradient echo T2* weighted: sensitive to iron or things that perturb the local magnetic field. The mainstay of functional brain imaging because it detects changes in the concentration of deoxyhemoglobin (based in the iron in the heme). Also used to quantify iron deposition in regions like the substantia nigra.

Spin echo, T1, T2 weighted: The workhorses of clinical MRI. The first types of MRI to be made.

Contrast enhanced images—adding contrast, usually gadolinium based that change the T1, to increase signal in specific locations. Often used to detect disruption of the blood brain barrier(tumors)

http://www.radiologyassistant.nl/en/p4556dea65db62/multiple-sclerosis.html    T1w MRI with Gadolinium enhancement to detect MS plaques

Angiography: using MRI to detect flow in the large vessels. This can be done with a T1w MRI and contrast. This can also be done using tricks controlling spins, without contrast using things like time of flight or phase sensitive angiography.

Perfusion imaging: Again this can be done with contrast and T1 imaging, or an elegant method called arterial spin labeling (ASL). The ASL image changes the spins in a specific plane of the body, often the neck and then detects the time needed for the label to reach the region of interest. This is rapidly growing in popularity for things that could impair blood flow.

Susceptibility weighted MRI or susceptibility MRI: these are sensitive methods to detect blood, or iron deposits. They will be used for liver iron overload, microbleeds, brain iron deposition.

Ultrashort TE or UTE imaging. The relaxation time T2 is fast. For normal tissues signal dies away with a ½ time of about 40ms. For cartilage and white matter it can be less than 5ms. Tissues with short T2 are hard to image because the signal is lost so fast. UTE imaging solves this problem and allows for beautiful images of things like ACL tears or cartilage injury.

Diffusion: water diffuses and so the proton may not be in the same place during the data collection as it was at the time of the first RF pulse. This can cause signal loss. Diffusion imaging greatly accentuates that signal loss and so is sensitive to water diffusion. The sequence uses gradients at different orientations to detect the direction the proton is diffusing. This was first used for conditions that cause swelling and edema, particularly stroke.

Diffusion tensor imaging and fiber tracking: This is new and generates fantastic looking images. The diffusion method was made much more precise for the direction of the spin. The idea is that water will diffuse along fiber tracts in brain or muscle and not across them. Computer programs have been developed to analyse the primary direction of diffusion, the tensor, in each point in the image and then determine if any adjacent point has a similar directional preference. It then draws lines connecting these similar points in the image, creating tracts that represent nerve fiber structure. This will be big in both brain injury and in studying developmental disorders.

Cine: images can be collected very fast, much line cinematography. These are used for imaging things like the beating heart. Fantastic.

This is a snapshot of image types. Higher fields and gradients are opening new opportunities. Better electronics and post-image processing methods are creating new types of images. As people figure out how best to manipulate those spins, they continue to develop new types of images that are sensitive to different aspects of disease.