Neuroimaging

Neuroimaging techniques

The aim of neuroimaging methods is to produce images of the brain or to visualize brain activity. The two major fields are structural and functional imaging. Many different methods are used for this purpose, including

• Electroencephalography (EEG)
• Magnetoencephalography (MEG)
• Positron emission tomography (PET)
• Magnetic resonance imaging (MRI)
• Functional magnetic resonance imaging (fMRI)

Structural neuroimaging

MRI

3T MRI device.

Protons ($H^{+}$)are particles with Spin $\scriptstyle {\frac {1}{2}}$. In a strong external Magnetic field $\scriptstyle {\vec {B}}_{0}$ they tend to line up with the field. The underlying effects can strictly be described only by the means of quantum mechanics. Nevertheless it is possible to imagine the mechanism via classical analogies.

The protons can be compared to a classical solid top. When a rotating solid top in a homogeneous gravitational field is not perfectly aligned with the vertical axis of the field, it starts a precession movement around this axis. Similarly the proton spins are preceding in the magnetic field with a characteristic frequency—the lamour frequency.

It is now possible to excite the proton spins by a RF pulse. Excited spins are not aligned with the external field any more, but are preceding in the xy-plane in a coherent way. This precession movement induces a measurable signal in a detection coil. When the RF pulse is stopped, 2 decay processes begin and the corresponding decay times are measured.

• Spin-lattice-relaxation ($T_{1}$), longitudinal relaxation: The spins realign with the external field.
• Spin-spin-relaxation ($T_{2}$), transverse relaxation: The spins precession movement is dephased and becomes incoherent.

Different types of tissue have influence on these relaxation times, producing contrast in pictures of different tissue.

Tomographic 3D-information is gathered by coding spatial information in gradient fields. These are overlayed on the external field and thereby modify the lamour frequency.

Diffusion-tensor MRI

(CC) Image: Thomas Schultz.
Diffusion tensor image of a human brain after white matter tractography

By applying a special gradient to the magnetic field the diffusion of water can be imaged. To get 3D diffusion information a series of measurements with different gradient directions has to be performed. From this diffusion data one can reconstruct the directions of fibers by a method called white matter tractography. This is necessary to create connectivity graphs of the brain.

Functional neuroimaging

Functional neuroimaging is used to image metabolic activity in neural tissue.

Positron emission tomography

A typical PET scanner device.

Radioactive labels are injected into the human body. In their decay they emit gamma-rays. These are measured and the source is localized by tomographic reconstruction.

Functional MRI (fMRI)

(PD) Image: National Institutes of Health
An fMRI study of patients with schizophrenia. While they performed a working memory task, the less the prefrontal cortex (red) activated, the more dopamine increased in the striatum (green).

Active regions in neural tissue show an increased blood flow. This results in a higher oxygenation of the blood in those regions. Because of the different magnetic susceptibility of oxygenated and deoxygenated blood this can be made visible via MRI methods. This technique is called blood-oxygen level dependent (BOLD) fMRI. It is often used in research to measure the reaction of the brain to certain stimuli or actions.

Data processing issues

Data provenance

To be useful as data resource in scientific collaborations, the history of images has to be tracked. This is called "provenance" (also "lineage" or "pedigree"). Provenance consists of the data part and the process part. The data provenance contains information about the origin of raw images, while the process provenance documents the way those images have been manipulated.

Brain atlases

For proper comparison between results of scientific studies one needs a method to locate e.g. areas of activity in a fMRI study. Brain atlases are a way to improve the comparability and indexing of study results.