My Brain
Most of my research starts with something relatively simple, a three-dimensional image made of numbers. Each number is the measured MRI signal at one small cube of space in the head called a voxel, the 3D counterpart of a pixel, typically about a millimetre across. A single scan contains several hundred thousand of these tiny blocks, each holding just one number. Put them together and process them with neuroimaging software, and those numbers become measurements of brain structure, connectivity, and development.
And below is a more concrete example: my brain! This is my own structural MRI scan, some of the raw material from which the measurements I work with are derived, rendered directly in your browser,
Click or drag to move the crosshair. Right-click and drag to adjust the contrast. Once you have clicked into the viewer, scroll moves through slices; until then scrolling just moves the page.
The brain is mine. The credit for making it goes to Mom and Dad.
What you are looking at
This is a T1-weighted anatomical MRI. It gives a detailed picture of the brain’s structure, how bright each tissue appears depends on how it interacts with the MRI signal. In a T1-weighted image, white matter appears bright because of its lipid-rich myelin, grey matter sits in intermediate shades, and water-rich CSF is dark. These contrast differences are what let us distinguish and measure the different tissues from a single image.
Around the outside are the scalp and skull. Just beneath them lies the cerebral cortex, the brain’s folded outer layer. It is made up mainly of grey matter, where nerve cells are densely packed, and is only about 2.5 mm thick. Beneath the cortex is white matter, made up of bundles of myelinated nerve fibres that carry signals between different brain regions.
The dark spaces inside the brain are the ventricles, which are filled with cerebrospinal fluid (CSF). CSF also flows around the brain and spinal cord, helping to cushion and protect them.
Deep in the centre sit the thalamus and, just in front and to either side, the basal ganglia: drag the crosshair into the middle of the axial view and they appear as paired grey masses, separated by a pale band of white matter, the internal capsule. Both of these are subcortical structures, literally, structures beneath the cortex. They also sit close to some of the questions I work on. The thalamus is the focus of one of my Generation R projects and has surfaced in my previous ENIGMA-OCD analyses; the basal ganglia anchor the fronto-striatal circuits that feature prominently in models of OCD.
From an images to measurements
The scan above is a structural MRI, a snapshot. What it can tell us is anatomical. Depending on the question, it can be processed to identify tissue classes, locate structures, and extract measures of their shape and size like cortical thickness, surface area, subcortical volume.
But a snapshot is only one way to look at a brain. Scan repeatedly while someone lies still and you get something closer to a film. Functional MRI (fMRI) tracks changes in blood flow and oxygenation over time, an indirect measure of activity that lets us follow how different regions behave, and how they behave together.
Diffusion MRI looks at the wiring instead. It measures how water moves through tissue, and because water moves more freely along nerve fibres than across them, that movement traces the white-matter tracts connecting one region to another, the brain’s structural connectivity.
Anatomy, activity, and wiring are three quite different views of the same brain. Multiply any of them by a few thousand people across a few dozen research sites, and you have the sort of dataset the research pages describe.
My scan has been defaced
A structural MRI can contain enough information to reconstruct a person’s face, so an unprocessed scan is potentially identifiable. This one has been defaced: facial features removed, brain untouched, which is common practice before sharing any structural scan, including your own.
The file has been resampled to 1.2 mm and quantised to 16 bits to keep the page light, so it is meant for exploration rather than quantitative analysis.
Rendered with NiiVue, an open-source WebGL viewer.