Radio sweat gland - 90 GHz

[Given that part of the point of this blog is to add a bit of value to stuff I publish elsewhere, I thought it was worth putting up this piece that appears this week in a necessarily abbreviated form in Nature's news pages. In particular, it's a shame not to hear from Merla, who is close to this topic, and more from the famous Paul Ekman.]


Sweating – a sign of recent physical activity and, often, of mental stress – can be detected from a distance by a beam of millimetre-wavelength radiation, a team in Israel claims[1].

They have shown that sweat ducts in human skin acts like an array of tiny antennas that pick up radiation at frequencies of about 100 gigahertz – the so-called extremely high frequency or EHF range, lying between microwaves and terahertz radiation. The antenna behaviour is all down to the ducts’ curious shape: they thread through the epidermis as regular helices. Filled with electrically conductive sweat, these channels act a little like coils of wire that absorb radiation across the millimetre and sub-millimetre wavelength band

Yuri Feldman of the Hebrew University of Jerusalem and his colleagues measured the reflection of EHF radiation from skin on the palms of subjects after 20 minutes of jogging, and found that there was a strong band of absorption compared to people who had not exercised. The absorption gradually disappeared as the jogging group rested. They also found that the reflection signals were proportional to blood pressure and pulse rate, known to be indicators of physiological stress.

And when the researchers suppressed sweating with a synthetic compound that mimics the localized paralysis of snake venom, inactivating the sweat glands, they found that the EHF absorption was lower.

Feldman and colleagues say that the helical antenna array makes skin a kind of biological metamaterial, in which the material’s response to electromagnetic radiation is determined by structure rather than composition. Metamaterials made from arrays of tiny electrical circuits are being explored for applications ranging from super-lenses to invisibility shields. “Nature has done what is being attempted extensively today in nanophotonics”, say the researchers.

Arcangelo Merla, who works on biomedical imaging of arousal states at the University ‘G. D’Annunzio’ in Chieti, Italy, calls the work “quite fascinating” and agrees that “it may open an alternative way for remote sensing of this important physiological phenomenon”.

They speculate that their technique could be used to gauge people’s mental state from a distance, perhaps even without their knowing. “This effect might be used for biomedical applications and homeland security applications”, the Israeli team say. Detection of sweating hands has previously been used as in lie detection, although the use of such physiological parameters in ‘polygraph’ lie detectors has become controversial after being strongly criticized in a 2002 report by the US National Academy of Sciences.

“Perspiration is related to increases in emotional arousal”, agrees Paul Ekman, a psychologist in Oakland, California, and one of the authors of that report. “But as with other measures of arousal, such as heart rate, it can be the consequence of many different mental processes. In terms of lying, arousal measures only tell you the person is aroused. Suppose you did not kill your spouse but the police are interrogating you: wouldn't you be aroused? The fear of being caught looks just like the fear of being disbelieved.”

Merla also points out that emotional sweating is driven and controlled in a different way from thermoregulatory sweating caused by exercise. He adds that “understanding a mental state from measures of peripheral activity is a very complex task”. He says that it would be inappropriate to apply the technique to lie detection and monitoring of stress and anxiety unless one combines it with other indicators of arousal. He is developing infrared thermal imaging of skin to determine several such measures simultaneously.

So far, however, Feldman and his colleagues are cautious about whether the idea will work at all, let alone how it might be applied. “We must first to evaluate the limits of performance – for example, what is the range at which we can detect a meaningful signal, how fast changes in the various biometrical parameters we want to monitor are manifested in our signal, and so on”, they say. “We are just starting our journey in these uncharted waters.”

Reference
1. Feldman, Y. et al. Phys. Rev. Lett. 100, 128102 (2008).