So why would anyone pay the extra million for MEG? That question was addressed by Gareth Barnes from University College London's Wellcome Trust Centre for Neuroimaging. Speaking at the IOP Medical Physics Group's recent meeting "Up and Coming Techniques in Medical Physics Translated into Clinical Practice", Barnes explained that the improved performance achievable by MEG is hard to ignore.

EEG results, he explained, are highly dependent on the properties of the patient's skull, making precise location of the source of internal electrical activity difficult to pinpoint. With MEG, however, the structure and composition of the skull have less impact on the results. MEG also offers direct, whole-brain imaging, and can measure electrical changes on a millisecond timescale. "We pay the extra million to say with confidence where the activity is," said Barnes.

He described the example of planning surgery for epilepsy, one of the main applications for MEG. In the 30% of epilepsy cases where patients do not respond to drug treatment, surgical removal of unhealthy tissue has proved effective. The use of MEG could allow more accurate localization of the seizure focus, particularly as this focus may propagate through the brain over time, a process that can be tracked using MEG's millisecond temporal resolution. But it is imperative to remove the correct regions of tissue. "The surgeon has to identify the unhealthy brain tissue to take out, and the healthy tissue to really not take out," noted Barnes.

To do this, surgeons often perform an additional operation to implant an electrode grid within the brain to measure a patient's seizures over a week. The use of MEG could remove the need for this additional surgical procedure. Indeed, Barnes noted that in cases where both techniques were used, MEG changed surgical decisions – whether to operate or not, or where to place the electrode grid – in about 20% of cases.

There are downsides, however. Whereas EEG systems are relatively low-cost, portable and can be worn, MEG scanners are expensive and not portable – and require the subject to sit perfectly still. The tiny fields being measured (10–12 to 10–15 T) also necessitate the use of liquid-helium-cooled superconducting quantum interference devices (SQUIDs) in a shielded room.

Barnes noted that the greatest clinical benefit of epilepsy surgery is in children. But as MEG systems are so expensive to manufacture, they are generally built to fit adult heads. When imaging child-sized heads, this results in a loss of sensitivity (as signal decreases with the square of distance) of a factor of two to four times. But there is a solution that can address all of these problems: get rid of the cryogenics.

Cryogen-free MEG

One way to ramp the sensitivity of a MEG system is simply to move the detectors closer to the subject's head. Simulations showed that in adults, this can provide a five- to 10-fold improvement in sensitivity. "And in young children, we expect 20 to 40 times better signal," Barnes added. This is not possible to achieve in SQUID-based MEG systems as the detectors' low temperature requires adequate insulation from the patient's scalp.

Another option lies in the optically pumped magnetometer (OPM), which can record biomagnetic signals without needing cryogens. Unfortunately, OPMs suffer from the opposite problem – they are based on in alkali atoms in the vapour phase, which requires heating of a cell of gas that still needs insulation from subject's head.

Recently, however, researchers have developed OPMs that are so small that very little heat dissipates outside the cell. These tiny detectors can be placed a few millimetres from the subject's head. Barnes cited a study using a 12 x 8 x 8 mm OPM that demonstrated comparable sensitivity to a SQUID-based system, albeit with higher noise level. He noted that last year the company QUSPIN was set up to commercialize such miniature OPMs

Barnes and colleagues Matt Brookes and Richard Bowtell at Nottingham University have just been funded by the Wellcome Trust to develop a wearable MEG system using these tiny OPMs. After determining the best placement for the sensors, they 3D printed a mask and helmet to hold 13 OPMs. They tested the set-up in a shielded room, applying electrical stimulation to the wrist and recording the generated signals in the cortex. "Compared with a SQUID, we saw a five-fold improvement in signal magnitude after stimulation," said Barnes.

Future prospects

So what does the future hold for cryogen-free MEG systems? For starters, Barnes predicts that the sensors will get smaller and cheaper (they currently cost about $8000). But there is still much development work needed. Currently, the sensors are highly sensitive to movement, which creates unwanted variation in the signal. The researchers propose that this can be addressed by the use of large coils to null the external fields, which could be built into the walls of the room, around the subject.

The ability for patients to move around in a wearable MEG helmet would open up many new experiments, new subject cohorts and clinical paradigms. Importantly, this approach could remove the need to implant EEG grids for surgical planning in children.

Barnes concluded that this new generation of cryogen-free sensors give a five- to 10-fold improvement in sensitivity and spatial resolution over current MEG systems. "Soon, there will be an affordable MEG sensor available that doesn't require maintenance and can be used in any hospital," he said.

Related stories

• SQUIDs line up for healthcare applications
• Algorithm cleans up biomagnetic data
• High-temperature SQUID outperforms rivals
• An easy-to-use foetal heartbeat monitor