Mar 27 2018

Wearable Magnetoencephalography

There are scientific advances, and then there are advances that help us make more scientific advances. The latter are often trickier to communicate to the public, because their connection to any tangible benefit is indirect. But improvements in research technology itself can have incredible potential to transform our future.

One category of scientific instruments that I am particularly interested in, as a neuroscientist, are ways to scan the brain, both anatomically and functionally. A functional scan is one that looks at how the brain is functioning in real time, it looks at the pattern of activity of brain cells. If we can correlate this brain activity with specific tasks, then that will teach us something about how the brain works.

Using this type of technology scientists are trying to reverse-engineer the brain, to map all of its connections (the connectome) and learn what the brain’s networks actually do. The ultimate goal is to be able to simulate those networks in a computer, or to build a computer that works like the brain. That would be the ultimate brain research tool – then you could run countless simulations and tests and see how the various networks in the brain behave.

As an aside, it’s likely that even a simulated brain, if it were functioning, would be self-aware. That raises a lot of ethical question in terms of further research.

Right now our tools for imaging brain function are low resolution. The “pixels”, if you will, are orders of magnitude larger than individual brain neurons. We can essentially see which part of the brain is active, and those parts are getting smaller, but we are nowhere near being able to image the activity of a single neuron in a living patient. Incremental advances in existing technology will get us pretty far, probably, but also probably we will need new technology to get the kind of resolution we need for neuron-level imaging.

One of the best existing technologies for imaging brain function is magnetoencephalography (MEG). This is similar to electroencephalography (EEG), except the latter measures electrical activity, while the former measures magnetic fields. Neurons conduct electricity – that is how they work. They are essentially biological electrical wires, and they communicate with each other through chemical signals (neurotransmitters).

And of course, electrical currents generate magnetic fields (and magnetic fields generate electrical currents). So, we can measure the magnetic fields generated by the electrical currents in the brain. We can also induce electrical currents in the brain with external magnetic fields (transcranial magnetic stimulation).

MEGs are massive machines because the magnetic fields generated by the brain are tiny. The MEGs have to be extremely sensitive. Current MEG machines are like giant hair dryers, with cooled superconducting magnets, and lots of shielding. Also, because the fields are so small, the subject cannot move at all, which limits their utility. So the subject is sitting perfectly still in a chair while the MEG is placed over their head (again, like an old-style hair dryer).

Therefore, any test involving movement is impossible. Researchers have had to simulate movement with virtual reality, but this has its limits. Also, subjects that cannot remain perfectly still cannot be studied – including children, the cognitively impaired, and patients with tremors or other movement disorders.

Here comes the incremental scientific advance – scientists have developed a new MEG technology that can be worn as a helmet. It uses a new kind of sensor:

To work around such workarounds, Bowtell’s team created a wearable 3D-printed mask that, instead of using superconductors as sensors, relies on 13 small glass cubes filled with vaporized rubidium. These optically pumped magnetometers (OPMs) get to work when a laser pulses through the vapor, lining up the atoms in its path. When neural current from the brain generates a small magnetic field, it knocks the atoms out of formation. A sensor on the other side measures fluctuations in the light from the laser to paint a map of brain activity.

Damn, we are clever chimps.

Right now these sensors are expensive, about $7000 each. The current helmet can only record from one area of the brain at a time. Building a more complete array of sensors could cost a million dollars, out of reach for most research labs. But of course this puts us on a new technological track, and hopefully further incremental improvements will increase the functionality of the helmet and bring down the cost. Smaller, better, cheaper.

One other limitation is that the device needs to be shielded from the Earth’s magnetic field. This requires a special room, so the device cannot be used out in the world.

This new MEG design allows for types of uses not possible with the static design. You can use it to study subjects while they engage in a physical activity, or in subjects with Parkinson’s disease or other movement disorders who can’t stay still. The team is designing a helmet for children and even infants.

This advance will produce a new round of information about how the brain works, and not just the healthy brain but what is different about the function of brains with specific diseases and disorders.

MEG mostly has a research application. It is used clinically, but its clinical use is very limited, mostly evaluating patients with epilepsy prior to getting surgery. The new device may have more clinical applications, but that remains to be seen. Perhaps it will have diagnostic uses for certain neurological disorders.

It’s always fun to look to the future with any technology, and to extrapolate from incremental advances. This is a particularly significant incremental advance, because it crosses a functional milestone – allowing the study of subjects that are moving. But where will this technology lead?

This specific technique will likely improve, but will also likely have limits. But scientists will find new ways to measure magnetic fields from the brain, ways that allow for smaller and more portable devices. The ultimate goal would be to have a slim cap or helmet, or even just small individual sensors, that can be attached rigidly to the scalp. All that is needed is for the sensors to not move with respect to the head. Ideally they would be shielded so that the subject can walk around in the world, go about their business, while their brain activity is being recorded in tiny detail.

And of course there may be ways of measuring and recording brain activity that do not use magnetic fields. Right now we have techniques for measuring electrical activity, blood flow, and metabolism to parts of the brain. These all have less resolution than MEG. But they are all non-invasive.

Invasive techniques, such as implanting nanowires in the brain to act as tiny electrodes, could theoretically have neuron-level resolution.

But of course the holy grail is a non-invasive technique that has neuron-level resolution in real time while allowing for full freedom of movement. We don’t know how this could be achieved, or even if it can (there may be ultimate limitations because of the skull and scalp getting in the way). But we are clever chimps, so it would not surprise me if eventually we figure out some way to do it.

Like this post? Share it!

No responses yet