How can the brain be mapped? - Chapter 4

How has brain imaging developed?

In the past, post-mortem studies were the only way to learn about the anatomy of the brain. The invention of the microscope and groundbreaking work by Golgi (who discovered the staining of tissue) created the idea that individual neurons exist. Ramón y Cajal (1889) then described that brain cells form independent units that communicate with each other. At the beginning of the twentieth century, Brodmann described the cytoarchitecture of the cerebral cortex. This ‘Brodmann atlas’ is still frequently used to identify cortical brain areas. Thanks to the imaging techniques that are available nowadays we have an even better view of the brain areas. In the past methods such as pneumoencephalography and encephalography were used. Pneumoencephalography is an invasive technique in which brain fluid is removed, and echo encephalography uses sound pulses. However, these are painful and / or do not provide clear images.

CT scan

‘Computed axial tomography’ (CAT) or computed tomography (CT) was introduced around 1970 by Cormack and Hounsfield. This technique is particularly suitable for providing a rapid assessment in the acute phase when a patient is admitted to the hospital (for example in the presence of a cerebral haemorrhage or skull fracture). However, CT scans do not show a clear picture of the different brain structures because the difference between gray and white matter is not depicted well: this makes MRI the preferred choice for scientific research.

The SPECT and PET techniques

In the 1980s it was proved that specific tissues in the brain can be stained by incorporate molecules with radioactive particles. The blood flow and metabolism within the brain tissues can be visualized with this. 'Single-photon emission computed tomography' (SPECT) or 'positron emission tomography' (PET) owe their success to this discovery. Both give an indication of the severity of the damage.

MRI, MEG and EROS

In the same period Mansfield and Lauterbur developed 'magnetic resonance imaging' (MRI). MRI measures changes in blood flow. MEG is magneto-encephalography. Magnetic fields are generated during neural activity. These are called event-related optical signals (EROS). Brain research can be done using infrared light and ultrasound.

What is structural imaging?

The CT scans and MRI scans are the most popular instruments for depicting the anatomy of the brain. Both methods complement each other. Imaging equals data collection; image processing is equivalent to 'image processing' of the acquired data.

Structural imaging with a CT scan

The CT scan uses X-rays that are passed through the body. Specific structures block more or less radiation. The result is a two-dimensional image. During a CT scan a lot of different images are made, because the x-ray source is constantly being moved. The images are then combined using a mathematical algorithm. The high resolution image offers the possibility to look into every slice of the brain. A slice can be sagittal, coronal or transverse.

• Sagittal: a vertical section from the nose to the back of the head, or from posterior to anterior: the brain is divided into a left and right hemisphere

• Coronal: a vertical section from bottom to top in the center of the brain, or from ventral to dorsal: the brain is divided into the anterior and posterior half.

• Transverse: a horizontal section from the eyes to the other side. Divides the brain into the dorsal and ventral part.

The bone is clearly visible and the ventricles are clearly distinguishable from the brain tissue, which means that the CT scan is often used for diseases such as schizophrenia. The CT scan is frequently used to see if there is a haemorrhage or a space-occupying lesion. The CT scan is used very frequently despite the lower spatial resolution than an MRI scan and the carcinogenic X-rays.

Structural imaging with an MRI scan

With an MRI scan, a clear distinction can be made between the gray matter (also called ‘neuropil’ consisting of the nuclei of neurons and other cells and deeper nuclei such as the hippocampus) and the white matter (or the wiring: the myelinated axons). The MRI provides insight into possible abnormalities in the white matter and is not harmful, in contrast to the CT scan. The MRI scan works via magnetic fields. A scanner with a field strength of 3 tesla gives a much sharper contrast image than a scanner with 1.5 tesla. Sometimes a contrast fluid is used to see brain damage more clearly. By using an MRI, it is possible to perceive the total volume of the brain or the volume of the brain parts in an accurate way. The thickness and surface area of ​​the cortex and the thickness and length of the white matter webs can also be derived with an MRI. With a T1-weighted scan of the entire head, the entire volume of the brain is visible. A T2 weighted image is created if white matter intensities are to be made visible.

Author's side comment: both a CT scan and an MRI scan provide a three-dimensional image of the brain. A CT scan uses X-rays while an MRI scan uses a magnetic field. A contrast agent can be added to both (with an MRI scan a different liquid is used than with the CT scan). People who have metal in their body such as pacemakers or insulin pumps cannot pass through an MRI scanner because the magnetic field will then be disrupted. With an MRI scan, the patient goes through a silent tunnel that unlocks the entire body (making an MRI almost impossible for people suffering from claustrophobia). The CT scan does not involve a tunnel, but a noisy ring through which the patient goes. With a CT scan the bones, brain, lungs and arteries / blood vessels are mapped. With an MRI scan, the brain and their function, the spinal cord, nerves, muscles, tendons, joints, and heart (function) can be distinguished.

What is structural image processing of the brain?

Image processing is about the processing and quantification of brain images. 

Volumetry and VBM

The total gray matter, the white matter and all the (cerebrospinal) fluid are added to the volume of the brain. Measuring the volumes and therefore the boundary between the gray and white matter is done by analyzing and labeling voxels (three-dimensional 'pixels' of slightly larger than 1 mm3). Nowadays this work is mainly done by using of computer programs. The average density (based on multiple scans of several people) of the gray and white matter is calculated per voxel by means of "voxel-based morphometry" (VBM). Only if the images of the scans are of very good quality the thickness of the gray matter can be derived from this.

Cortical thickness

Cortical thickness involves studying the thickness of the cortex at a certain place. This differs from the above-mentioned techniques. You can also look at cortical surface. The difference is that thickness says something about the number of cells in a column, and the cortical surface says something about the number of columns.

DTI

Diffusion tensor imaging (DTI) scans utilizes the properties of water molecules that move freely in all directions (isotropic diffusion) or limited movement (anisotropic diffusion). The DTI provides information about the direction of the white matter in the brain.

MRS

Magnetic resonance spectroscopy (MRS) provides information about the concentration of certain molecules. This technique is frequently used to examine a specific location in the brain in detail.

What is functional imaging?

Just to measuring the amount of metabolism of a brain structure to see which area is active is not enough; a resting area can have a higher metabolism than an 'active' brain part. For this reason, a method is required that can measure the change in brain metabolism. The structural imaging only makes 3D "images", while a functional imaging creates 3D "films" (several photos in succession). The techniques differ from each other in both spatial resolution (the sharpness of the images) and the temporal resolution (the speed of the recordings). An EEG has the best temporal resolution, followed by fMRI and in a shared third place the EEG and MEG. The PET scan has the best spatial resolution, closely followed by the MEG, PET and fMRI.

How are electrical signals measured?

Action potentials create electric fields that can be measured on the outside of the head. There are two types of electrical activity: within a brain region, and communication between different brain regions. The functional imaging method that is used most commonly is electroencephalography (EEG). Disadvantages of EEG are the disruptance of the electric fields by the skull, the skin and the surrounding tissue. The result is a fairly large error margin. Magnetoencephalography (MEG) measures the magnetic fields generated by action potentials. MEG is not disrupted by tissue or bone. This enables a more precise location of activity.

Eelectro-encephalography (EEG)

EEG is the oldest and simplest functional imaging method. With an EEG, electrodes are placed on the skull and (after many scans) three types of information can be retrieved:

1. The event-related potential or ERP is the average electrical signal from the tissue closest to the electrode. This signal provides information about the course of time and the strength of the neural response.

2. Fluctuation of the EEG signal is the result of brain waves. Specific patterns (such as alpha waves) are associated with certain brain activity. The amplitude of the waves can be calculated to make an estimate of, for example, the concentration.

3. The locations of the areas that respond to a specific task can be determined by the relative strength on each electrode.

An EEG is not sensitive enough for individual neurons, but it does have a large temporal resolution.

Magneto-encephalography (MEG)

MEG measures with a smaller error margin and also measures brain waves that cannot be seen with an EEG: the high gamma waves. The MEG scan is very expensive and requires a lot of maintenance. MEG offers the same types of information as EEG, but the calculations are more complex. In contrast to EEG, the accuracy and sensitivity are much less because the deeper areas are analyzed.

What are haemodynamic signals?

PET was the first method to measure neural activity through haemodynamic signals: water was made radioactive using a radioisotope, and added to the bloodstream. The extent to which water is absorbed by brain tissue is dependent on local oxygen consumption: active areas need more oxygen. A major disadvantage of the PET is the radiation damage. For this reason, fMRI is preferred.

fMRI

Functional MRI uses an MRI scanner. Due to the slow haemodynamic response compared to the neural response itself, the temporal resolution is not optimal. The spatial response is very good. The assumption is that haemoglobin (Hb) is used in the blood as a natural contrast fluid. The Hb molecule is responsible for the transport of oxygen to the body tissues. Where neurons are more active, the blood supply increases and more oxygen is taken from the Hb molecules.

What is functional image processing?

There are two relatively new methods for analyzing functional imaging data. In these cases, we do not look at each area separately, but at the coherence between the brain areas. These methods are called the "resting-state" method (where the person relaxes for 5 to 10 minutes) and the "mind reading" method (where visual stimuli are offered).

How are the mentioned techniques useful?

Most techniques are useful for research into abnormal structures and functions in neurological and psychiatric disorders. It is also used to perform better differential diagnostics.