Struppa: Measuring the seemingly immeasurable
May 12th, 2008, 2:00 am · Post a Comment · posted by grobbins
In his latest column, Chapman University Chancellor Daniele Struppa discusses advances that allow researchers to cut through the “noise” of science to find, measure and exploit important data.
You arrive at the party, and as you enter the room you see lots of happy people talking, laughing loudly, and, to top it off, there is a band in the back of the room, playing some old tunes from the fifties. As you look around, you see an elegant young man whispering something to your wife. … Wouldn’t you wish you could pick up exactly what he is saying (and her response) instead of all the chatter and music?
This scenario is something that scientists face in a wide variety of contexts. For example, biomedical researchers are now trying to isolate, in the blood of potential patients, the proteins (biomarkers, as they call them) that may indicate the incipient stage of a cancer. However, in many cases, such proteins are very small and their presence is hidden by large quantities of large and (at least from this perspective) unimportant proteins (for example, albumin, which we all have in large quantities in our blood).
How can we isolate the small protein we are interested in, instead of being distracted by the irrelevant proteins? A terrorist may have introduced a small (and yet lethal) quantity of a biotoxin in the air or in our water reservoir. How can we detect such traces of biological materials, when both the air and the water are full of other such materials, whose presence overwhelms the biotoxins we want to identify? And finally, more to the original example, how can we detect an acoustic signal (maybe from an enemy submarine, maybe from a terrorist conversation) in an ocean of noise?
There are at least two technical issues to deal with in these scenarios. One has to do with what engineers call “denoising,” i.e., the development of techniques used to remove noise from a signal. This is something we all are very aware of. For example, when we talk to a friend, our brain does a lot of automatic denoising. The noise from the road, from the television in the background, and from other sounds is filtered out, so that we can concentrate fully on our conversation (at least as long as the noise is not too much). We only become aware of noise when denoising techniques do not work too well (for example, when we are on our cell phone, and the static becomes an impediment to the conversation), or when the noise is so strong that it cannot be reasonably eliminated. The past 20 years have seen great new developments in denoising (both for audio and video signals), and while this is a topic of great interest and relevance, we will probably put it aside for another column.
What I want to discuss here instead, is the second technical issue which is involved in the examples I have given before. Namely, even if we were to remove all the noise, how can we measure something which is so small that it falls below the capability of current measurement techniques. For example, how do we detect a contaminant, which is present in such minimal quantities as to be virtually invisible? How do we detect a biomarker in the blood, when its presence reaches levels below the sensitivity of our instruments?
The recent Feb. 8 issue of Science seems to indicate that a new revolutionary step has been taken in this direction. A recent experiment of O. Hosten and P. Kwiat indicates that it is in fact possible to amplify signals up to 10,000 times, so that the phenomenon that they describe can now be measured by standard instruments. The specific experiment which is described in the two articles in Science has to do with the deviation of a ray of light. In a lab experiment, the two scientists from the University of Illinois at Urbana-Champaign were able to detect a deviation of only a few nanometers (one nanometer is only one billionth of a meter and, to give an idea, a typical germ measures about one micron or 1,000 nanometers).
This is particularly impressive because the vibration of the room where the experiment is conducted is, by itself, sufficient to hide such a shift. In addition, the usual instruments can only detect variations of the order of microns (a micron is equivalent to 1,000 nanometers), and therefore such a feat seemed impossible to achieve.
What is even more fascinating is the fact that the technique used by Hosten and Kwiat is a first stunning application of a theoretical result first discovered, exactly 20 ago, by Professor Yakir Aharonov (one of the great quantum physicists of the last century, slated to join the faculty at Chapman University this coming academic year). In an article in Physics Review Letters, Aharonov and two of his collaborators (David Albert and Lev Vaidman) were the first to introduce the notion of ‘weak measurement’. This is a sophisticated idea which exploits the strangeness of quantum physics to allow the measurement of quantities that would otherwise be undetectable.
For many years, the concept of “weak measurement” has been rather controversial, and it was only a few years ago that a team from Rice University utilized this notion to slightly amplify a signal (about 30 times). This result, however, only showed that Aharonov’s idea could be applied, but was not enough to offer any practical application. The new discovery of Hosten and Kwiat, on the other hand, managed to take advantage of the notion of weak measurement to truly open a new vista on the field of metrology (that area of physics and engineering that deals with improving the way in which we measure lengths, weights, volumes, etc.).
The lag between the 1988 paper of Aharonov, and the 2008 paper of Hosten and Kwiat is not unusual in science, and it reminds us of the importance of fundamental research. Even results which appear to be esoteric, and possibly simply of theoretical interest, may (and often do) turn out to be central to important technological advances.
