The brain is one of our most essential organs. It holds a lot of information including knowledge, thoughts, and memory. That’s 100 billion neurons worth of information. Except, we barely know anything about those neurons or how they interact in the brain. That’s not to say we’re not trying to figure it out. It’s currently happening through connectomics.
You Haven’t Heard of Connectomics?
Connectomics is the study of the human connectome. The human connectome? Essentially the human connectome is a 3D structure that shows all of the neural pathway connections in the brain. The brain is made up of 100 trillion connections. That’s insane. To give you some perspective, the universe is only made up of 100 billion stars.
Functional connectomics is a subset of connectomics focusing on the collective set of functional connections in the brain. A functional connectome has the purpose of measuring the connectivity between particular brain regions.
These correlations are measured through acitivity over time between brain regions. Functional connectomics is currently our mainstay of research into brain networks. Correlation-based functional connectivity explorations as well as data-driven approaches such as independent component analysis and machine learning are being used to analyse functional connectomes.
Breakthroughs in Connectomics
It terms of connectomes, we’ve only fully mapped the connectome of the roundworm C. elegans and Google’s working on developing a high resolution connectome of the fly brain.
How’re We Measuring This Data?
Functional connectomics is measured at a macroscopic scale providing researchers with millimetre resolution through the use of non-invasive neuroimaging methods.
Functional connectivity is the statistical relationship between specific physiological signals in time through techniques such as functional magnetic resonance imaging (fMRI), electroencephalography (EEG) or magnetic electroencephalography (MEG).
Resting state fMRI is a relatively recent development that allows investigators to explore the modular nature of cortical function to assess resting sate functional connectivity. Functional connectivity can be measured using resting-state functional MRI (rs-fMRI) which measures the blood oxygenation level dependent signal when subjects are positioned in the scanner awake but are not performing a task.
The Power of fMRI
Not only can fMRI help diagnose diseases in the brain, but it might also enable doctors to get inside our mental processes to determine what we’re thinking and feeling.
fMRI is a non invasive technique that uses strong magnetic fields and radio waves to created detailed images of the body. Instead of creating images of organs and tissues like MRI, fMRI looks at blood flow in the brain to detect areas of activity.
fMRI may also be used to examine the brain’s functional anatomy, (determine which parts of the brain are handling critical functions), evaluate the effects of stroke or other disease, or to guide brain treatment. fMRI may also detect abnormalities within the brain that cannot be found with other imaging techniques.
MRI vs. fMRI
fMRI scans use the same basic principles of atomic physics as MRI scans, but MRI scans image anatomical structure whereas fMRI images metabloic function. Thus, the images generated by MRI scans are like three dimensional pictures of anatomical structure.
The big advantage of fMRI is that it doesn’t use radiation like X-rays, computed tomography (CT) and positron emission tomography (PET) scans. If done correctly, fMRI has virtually no risks. It can evaluate brain function safely, non invasively and effectively.
BOLD Contrast Imaging
fMRI uses Blood-oxygen-level-dependent imaging, or BOLD-contrast imaging to observe different areas of the brain or other organs, which are found to be active at any given time. BOLD contrast derives from variations in the magnetic susceptibility of blood due to variations in the concentration of deoxyhemoglobin.
Analyzing Functional Connectivity
There’s a few ways we’re currently using fMRI and other scans to analyze functional connectivity in the brain.
1. Seed-Voxel Correlation Mapping
Seed-voxel correlation mapping is the simplest technique for studying functional connectivity. The correlation coefficient between the fMRI signal at different times and measurements of the activation in a region is calculated for each voxel in the brain. This data is then displayed as a parametric image.
Seed-voxel mapping is usually applied to time series data where the experimental condition doesn’t change during an imaging run. Connectivity maps made using one seed region can be used to identify other regions to be used as subsequent seeds, after which the entire process can be started again. This approach can help reveal networks of functionally connected regions that otherwise remain unidentifiable.
2. Structural Equation Modelling
Structural equation modelling (SEM) essentially takes a step beyond calculating correlations. SEM is calculated via a maximum likeliehood procedure which adjusts the parameters of the model until the predicted correlations match the correlations in the data as closely as possible. The connection strengths are used to compare experimental groups or conditions.
SEM has been used with fMRI data from experiments including studies of visual attention, visual learning, grammar learning, tone listening, semantic and episodic memory, working memory, reading, and finger movement. The weaknesses of the approach come from the failure of the model accounting for time dependance of fMRI data.
3. Psychophysiological Interaction
A psychophysiological interaction (PPI) technique used to determine a stimulus or context dependent change in the influence of one brain region on another. A PPI uses a linear regression model to predicte influence on a region’s data, such as stimulus related changes. They allow for a whole brain search for voxels that exhibit response to a particular influencing region.
PPIs have been used to identify abnormal changes of anterior cingulate connectivity during language experiments in schizophrenic patients and to examine the role of the amygdala in response to fear signals.
4. Principal Components Analysis
Principal components analysis (PCA) decomposes imaging data into a series of components. PCA is applied where the data is repeated fMRI scans during an experimental run. The spatial variation in the signal is represented by an image called the eigenimage and the temporal variation is represented by the eigenvector. The eigenimages capture features of the data that are expressed in multiple regions of functional connectivity.
PCA has been used with fMRI for a variety of investigations, including to differentiate visual processing and semantic analysis, to study interacting areas responsible for the processing of metaphorical speech, and to classify networks responsible for different aspects of visuospatial processing. The disadvantages of the approach is the difficulty of interpreting the many eigenimage pairs that can be produced from fMRI data.
5. Partial Leas Squares
Partial least squares (PLS) is mathematically very similar to PCA, as it uses a correlation matrix. But instead of looking at the covariance of the data, PLS examines the correlations of image data values with some other variables. PLS produces images and associated vectors of scores for each image to understand connectivity.
PLS was introduced to neuroimaging in the context of PET rCBF brain imaging. PLS has been used for PET functional connectivity studies of episodic memory, sensory learning, and amygdala function. A spatiotemporal variant of PLS has been developed for fMRI time series data and applied to studies of autobiographical memory.
6. Multivariate Autoregressive Models
Multivariate autoregressive models (MAR models) incorporate the cross-covariances between regions and understand the temporal relationships between different scans to produce conclusions about the predominant directions of influence between regions as well as their strength.
- Connectomics is the study of all of the neural pathway structures in the brain.
- Functional connectomics focuses on understanding the connectivity between specific brain regions.
- fMRI is the most commonly used neuroimaging technique using Blood-oxygen-level-dependent imaging to observe activiy different areas of the brain.
- There are six common ways to analyze functional connectivity, through seed-voxel correlation mapping, structural equation modelling, psychophysiological interaction, principal components analysis, partial least squares, and multivariate autoregressive models.