Static MRI Contrasts & Pulse Sequences
One of the primary benefits of MRI over other imaging modalities is the wide range of tissue that can be nicely visualized using one of many contrast mechanisms (contrast being the signal difference between any two types of tissue). There are two overarching types of contrast mechanisms for MRI, static and motion. Static contrast can reveal tissue proton density, relaxation time after energy excitation, chemical concentration (ex. acetylcholine), and molecular type; motion contrasts provide information on blood flower, water diffusion and capillary irrigation. For functional MRI studies, the contrast used is most always the deoxygenated hemoglobin in the brain region of interest.
To understand how these various contrast methods work, one must first be familiar with the conceptions of repetition time and echo time. Repetition time (TR) is the interval between successive excitation pulses from the radiofrequency coils, while echo time (TE) is the time between the excitation pulse and data acquisition from the center of k-space. These values differ across tissue types and magnetic field strengths, and, more importantly, they can affect the degree of T1 recovery and T2 decay that is obtained. Generally, T1 recovery can be minimized with a very short or very long TR, while T2 decay can be minimized with a very short or very long TE.
The simplest type of contrast in MRI is the proton density static contrast, which creates images based on the number of protons in any given voxel (unit of space, in MR vocabulary). It is completed by running the MR scan with a very short echo time (TE) and a very long repetition time (TR), such that both T1 recovery and T2 decay are minimized. An unfortunate result of the long TR time is that the collection of images takes a similarly long time, but by reducing the flip angle to below 90-degrees, shorter collection times can be maintained.
In performing a proton density contrast study, either a gradient-echo (GRE) or spin-echo (SE) image sequence can be used. In a GRE image, a series of gradients alone is used to create the signal at the center of k-space, whereas SE utilizes a 180-degree electromagnetic pulse--the refocusing pulse--to do so. The GRE approach is more common with proton density studies because it is fast, accurate, and provides levels of intensity on the image (aka amount of brightness) that are directly proportional to the density of the underlying tissue. As such, one can differentiate fluid, grey matter and white matter. This segmentation of tissue type can be even further enhanced, though, by acquiring T1 and T2 weighted images as well.
A T1 contrast image is obtained via using either a spin-echo or gradient-echo sequence, and an intermediate TR with a very short TE; in this way, longitudinal recovery will be similar across tissue types, allowing intrinsically different tissue responses to T1 recovery to be seen. Typically, T1 is used to visualize white matter and bone marrow (both of which have short T1 values), while fluids become black and lose detail. There is a technique known as inversion recovery that allows enhanced T1 contrast by adding a 180-degree inversion pulse before the regular pulse sequence, but the details of this are beyond the scope of this discussion.
A T2 contrast image requires a spin-echo sequence with an intermediate TE time and a very long TR time. This allows for total T1 recovery with the largest variance in transverse magnetization between tissues. T2 images tend to make brighter those tissues that are more homogenous; as such, fluid (blood, CSF) is bright on T2, followed by gray matter, and then white matter is darkest.
A third type of contrast, T2*, is perhaps most important for functional neuroimaging, as it underlies most modern fMRI sequences. In T2*, a long TR and average TE value are used with a gradient-echo sequence to allow sensitivity of the image to local magnetic field inhomogeneities. These inhomogeneities allow visualization of deoxygenated hemoglobin in the blood, particularly the venous system, and thus form the basis for BOLD (blood oxygenation level dependent) fMRI imaging.
Another type of contrast--chemical contrast--utilizes the variance in resonance frequencies of individual protons that arises due to shielding from the surrounding molecule. Each molecule, then, has a distinct set of resonance frequencies, and by decoding the MR signal at predefined frequencies for particular molecules, we can resolve both the location and concentration of molecules of interest, such as glucose or creatine. This is, however, an inordinately time-consuming process.
A similar process, magnetization transfer, allows for the visualization of particular macromolecular regions of the brain, such as myelin (useful in, say, multiple sclerosis), by measuring the influence of macromolecules on the magnetization of protons in nearby water pockets. It works by emitting an energy pulse at a particular frequency sufficient to energize the macromolecule in question but not the surrounding water molecules. With time, some of the magnetization of the macromolecules will passively transfer to the water, after which a second energy pulse will be delivered and the difference in signal between the two pulses determined. The amount of magnetization transfer depends directly on the concentration of the macromolecule in question, and thus this procedure can give very good data on the location, amount, and quality of certain compounds.
Reference: Huettel, S. A., Song, A. W., & McCarthy, G. (2008). Functional magnetic resonance imaging (2nd ed.). Sunderland, Mass.: Sinauer Associates.
