Can a 'microwave helmet' really detect strokes?

“Microwave helmet ‘can spot a stroke’,” reports BBC News.

There are two types of stroke. The majority of strokes are caused by a clot stopping blood flow to an area of the brain. This type of stroke can be treated with anti-clotting medications to break up or dissolve blood clots. However, this type of treatment is disastrous if the stoke turns out to have been caused by bleeding into the brain.

Currently the only way to tell the difference is for a patient to have a scan in hospital. Going to hospital and waiting for a scan can delay treatment, and the sooner treatment is given, the less damage the stroke is likely to do.

The headline on the BBC was prompted by a proof-of-concept study that has shown that a “microwave scattering” technique can distinguish between the two types of strokes. The helmet device used by researchers is portable and could therefore be used by paramedics and other health professionals before a patient gets to hospital. This could allow treatment to start vital minutes earlier.

In the studies, when the cut-off was set to identify all haemorrhagic strokes, some people with ischaemic strokes were misclassified. But the researchers hope that information from a larger data-set from an ongoing clinical study will allow them to better differentiate between the two.

This early stage research is encouraging, but further work is required before NHS ambulances are equipped with “microwave helmets” for people who may have had strokes.

Where did the story come from?

The study was carried out by researchers from Chalmers University of Technology, University of Gothenburg, Sahlgrenska University Hospital and MedTechWest, all in Gothenburg, Sweden.

It was funded by VINNOVA (Swedish Government Agency of Innovation Systems) within the VINN Excellence Centre Chase, by SSF (Swedish Foundation for Strategic Research) within the Strategic Research Centre Charmant, and by the Swedish Research Council. It was published in the peer-reviewed journal IEEE Transactions on Biomedical Engineering.

The research was well reported by the BBC.

What kind of research was this?

This study was a description of the background, design and signal analysis of two microwave-based stroke detection systems, and a proof-of-concept clinical study on people who’d had a stroke and healthy people.

The researchers wanted to develop a new way of diagnosing stroke that was capable of differentiating between ischaemic stroke (caused by clots stopping blood getting to the brain) and haemorrhagic stroke (caused by bleeding into the brain). They wanted it to be used when patients arrive at A&E or by paramedics to enable appropriate anti-clotting medication to be started as soon as possible in people who have a stroke caused by a blood clot (ischaemic stroke). The need to distinguish between the two types of stroke is vital, as giving anti-clotting treatment to someone with a haemorrhagic stroke could be disastrous.

The researchers were interested in applying “microwave scattering” to this problem. They developed two prototype helmets with 10 or 12 microwave patch antennas. One at a time, each antenna is used as a transmitter, with the remaining antennas in receiving mode.

Microwave scattering can detect strokes because the scattering properties of white and grey matter are different to those of blood. The power output of the imaging systems were about 1mW, about 100 times lower than the 125mW transmitted by mobile phones.

What did the research involve?

In the first clinical study, 20 patients diagnosed with acute stroke were studied in a specialist hospital clinic between seven and 132 hours after stroke onset. Of the 20 patients, nine had a haemorrhagic stroke and 11 had an ischaemic stroke. In this study the first prototype was used, which was based on a bicycle helmet and had 10 patch antennas.

In the second clinical study, 25 patients with stroke were studied on a hospital ward, between four and 27 hours after stroke onset. Of the 25 patients, 10 had a haemorrhagic stroke and 15 had an ischaemic stroke. In addition, 65 healthy people were imaged. In this study the second prototype was used, which was a custom-built helmet with 12 patch antennas.

The signals obtained were analysed by a computer algorithm.

What were the basic results?

In the first clinical study, if the cut-off was set to identify all patients with haemorrhagic stroke, four of the 11 patients with ischaemic stroke were misclassified with haemorrhagic stroke.

In the second clinical study, when the cut-off was set to identify all patients with haemorrhagic stroke, one of the 15 patients with ischaemic stroke was misclassified with haemorrhagic stroke.
The technique was even better at distinguishing between patients with haemorrhagic stroke and healthy people.

How did the researchers interpret the results?

The researchers conclude that: “The relative simplicity and size of microwave-based systems compared to CT [computed tomography] or MRI [magnetic resonance imaging] scanners make them easily applicable in a pre-hospital setting. We suggest that microwave technology could result in a substantial increase of patients reaching a stroke diagnosis in time for introduction of thrombolytic [anti-clotting] treatment.

“The socioeconomic ramifications of such a development are obvious not only in the industrial world but also, and perhaps even more so, in the developing world,” they said.

Conclusion

This study has shown that haemorrhagic strokes could potentially be distinguished from ischaemic strokes by analysing microwave scattering measurements.

While the two types of stroke can already be accurately diagnosed by CT or MRI scans in hospital, the “microwave helmet” development is important because it could potentially be used before someone arrives in hospital. This would avoid any time delay and allow people with ischaemic stroke to receive the anti-clotting medication that they need as soon as possible, potentially reducing the extent of damage the stroke causes.

The technique isn’t perfect yet, but the researchers are hopeful that information from a larger data-set from an ongoing clinical study will improve the predictive power of the algorithms.

They also say that “introduction of pre-hospital thrombolytic [anti-clotting] treatment based on a microwave scan diagnosis will have to await studies of larger clinical cohorts”.

So while this early stage research is encouraging, further work is required before “microwave helmets” are used to distinguish between ischaemic and haemorrhagic strokes. More work is also needed to prove whether they could improve the care and treatment of people who’ve had a stroke.