Recent advances have led to an increase in the spatio-temporal resolution of techniques such as functional magnetic resonance imaging (fMRI). These advances mean that it will be possible, in the not-too-distant future, to observe the activity of small groups of neurons, or perhaps individual cells, in real time.
Diffusion MRI is usually used to observe the diffusion of water molecules in the brain, and makes possible the mapping of large fibre tracts and their connections.
Le Bihan, et al have used diffusion MRI to observe the activation of visual cortical neurons. They assume that changes in the movements of water molecules, which produce cell swelling and membrane expansion, are among the earliest biophysical events that take place in activated cells.
When firing, neurons require oxygen, which is supplied in the blood, and conventional fMRI measures the magnetic signals produced by the movement of oxygenated blood (hence its name, blood-oxygen-level, or BOLD, fMRI). As cells are activated, their supply of blood increases, and it is this increase that is detected by BOLD fMRI.
The biophysical changes observed by Le Bihane and his colleagues precede the vascular response detected with BOLD fMRI by several seconds, therefore providing more direct observation of neuronal activation.
Although it is not yet clear exactly what the signal produced by diffusion MRI represents, the method provides much better temporal resolution than BOLD fMRI.
New imaging technologies already in use at Massachusetts General Hospital (MGH) are providing increased spatial resolution to fMRI. Whereas traditional MRI uses a large magnet coupled with a radio-frequency detector to generate images of the brain, researchers are developing devices containing arrays of between 8 and 256 detectors to produce more detailed images. The signal-to-noise ratio increases with the number of detectors used.
Each detector is placed over a specific part of the brain to provide a high-resolution image of that part, and the images obtained from each detector are then brought together to produce a composite image of the entire brain.
An fMRI machine with an array of 32 detectors is already available from Siemens. In collaboration with that company, researchers are developing denser arrays of smaller detectors; the smaller the detector, the higher is the resolution of the image it produces.
The devices being developed will eventually reach a resolution of 10 micrometres, the diameter of an average neuronal cell body, and the next generation of neuro-imaging devices may be integrated with techniques such as molecular imaging and electroencephalography. Such machines will be very useful for research into a wide variety of neurological disorders.