March 11, 2006

Have you ever been stuck in a traffic jam on a June afternoon in Lahore? You are at the head of the lane and the traffic light is red. You are listening to your prospective employer on the cellphone, who is trying to negotiate a lucrative offer, but the cellular company is known for its poor signal coverage and you are having difficulty in making sense of what he is saying.

All of a sudden there is a cacophony of horns and loud discouraging shots, reminding you that the traffic lights have switched colours. Your employer’s speech virtually turns into a murmur dying out in the auditory tumult of the environment.

Observing nuclear spins in a magnetic resonance experiment is much like listening to the murmurs of your employer against the huge roar of horns and traffic. Many of us enjoying the luxuries of urban life can relate to an MRI scan directly or indirectly.

The abbreviation stands for “magnetic resonance imaging”. The scan is simply a snapshot of nuclear spins placed in a strong magnet, and the spins are emanating, not sound, but electromagnetic radiation. The radiation lies in the hundreds of Megahertz range and can be picked up with a properly designed radio that has been tuned in to receive such high radio frequencies.

Strictly speaking, the radiation emitted comprises several frequencies which are characteristic of the kinds of spins present and their environment inside the molecule. By looking at an imprint of the frequencies, it is possible to make intelligent guesses of the molecule’s internal structure. But the intrinsic problem with MRI is its low sensitivity — the signal originating from a nuclear spin is very weak.

In the context of my analogy of the traffic jam, the signal is hardly audible and is imbedded in a loud and mainly random clutter of background noise. In fact, present day technology falls short of making sense of the radiation from a single nuclear spin, which is simply too weak to be untangled from the noise. What then, makes an MRI scan possible?

To answer this question and to tackle the problem of insensitivity, one needs a foray into quantum statistical mechanics: quantum mechanics is one of the most powerful theories of physics that is applied to a collection of a large number of particles, in our case nuclei.

A select class of nuclei such as hydrogen, carbon, nitrogen, phosphorus and fluorine possess a property called the ‘spin’. Like many quantum properties, the spin is difficult to interpret but its manifestations are evident. The spin translates into the nuclei behaving as tiny magnets (about 1027 or billions of times smaller than the magnet inside your cellphone’s speaker).

Furthermore, in liquid state, these tiny magnets, attached to molecules, are whimsically reorienting in all random directions. The net effect is, therefore, that the liquid sample is “unmagnetised”.

However, we can place the sample inside another extremely strong magnet. In the presence of the extra magnet, the spin magnets tend to line up parallel or anti-parallel to the direction of the field.

The parallel and anti-parallel directions may be called the “up” and “down” states. The spin has lower energy when it is in the up state resulting in a very minute excess of spins in one state over the other. In typical conditions, for 500,000 down spins there will exist 500,001 up spins, the fractional population difference being only one in a million. But no matter how small it is, this net excess must exist to produce an MRI signal.

An excess of a single spin is insufficient to elicit a detectable signal, but if we have a bigger volume of sample and hence a large number of excess spins, we can easily cross the detection threshold. This is the situation when the spins are all acting additively and contributing to the strength of the signal. The individual spins are not louder, we just have more of them.

However, it is not all that simple. The up and down states are separated by an energy gap. Temperature is providing the spins with energy kicks strong enough to make them step over the gap and fluctuate between the two orthogonal states. Resultantly, this temperature-dependent energy tries to randomise the spins and counter act the effect of the applied magnetic field. The result is an ongoing competition between magnetic ordering and thermal disordering. At the end, some sort of compromise is reached and the spins find themselves happy with a fractional population difference that is determined by two factors, the temperature and the size of the energy gap.

Apparently, the population difference (also called polarisation) can be enhanced by lowering the temperature and making the spins less agitated; or by increasing the applied field and widening the energy gap which thereby becomes more difficult to cross.

How far can these approaches take us? Not too far. In a conventional magnetic image of the human tissue, we are rescued by the profusely large number of spins, namely hydrogen nuclei or protons, thanks to the abundant water, protein and lipid contents of these tissues.

However, nature is not always so generous and many important chemicals only exist in trace amounts. Think neurotransmitters serotonin and dopamine, which control our innate feelings of excitement and fear, anguish and delight, love and rage. These signalling chemicals, vital they may be, come in exceedingly small concentrations and the MRI signal they produce will be exceedingly small. Furthermore, it is not possible to artificially increase their concentration in the naturally available setting.

Another option is to lower the temperature and try to freeze the spins and locking them in the up state. But this approach has its own problems. The energy gap is so small that even very cold temperatures can randomise the spins. Thus, for achieving significant polarisations, one has to move down not just to the cold, but to the extremely cold, for example, 260 degrees below freezing! This is even colder than the surface of the planet Pluto. These temperatures are surely achievable but impractical for most cases of interest. Quite obviously, the human subject cannot be subject to such inhuman temperature zones and for non-medical applications too, most liquids will freeze well above this temperature.

The third option for increasing the spin polarisation is to increase the magnetic field. Unfortunately here, we are limited by technology. The strong magnet in an MRI experiment is both similar and different to the solenoid magnet we are accustomed to, for example, in our doorbells. The magnetic field is produced by passing current through a coil of wire, as in the doorbell solenoid, but the material used is not copper but a super-conducting alloy of niobium and tin. The zero resistance of the material allows the coil to carry an exuberant amount of current and that too without a power source or dissipation of heat. The large current produces intense magnetic fields, between 1 and 20 Teslas.

However, there are technological limits to the maximum field achievable: we are limited by the maximum current the alloy can carry without losing its super-conducting properties and structural stability. The current record is above a Kilo Tesla in an explosive underground, quarantined experiment, where neither the magnet nor the sample survived. Saner experiments set the current limit to only 20 about Teslas where the enhancement in spin polarisation is insignificant. Furthermore, one of the very reasons why MRI is so rarely administered in many developing countries is the high cost.

On the outset, the situation does not look promising. However, the antidote to many a poison lies in the poison itself. Ironically, quantum statistical mechanics also presents one way to overcome the polarisation problem it had created in the first place.

The hydrogen molecule, comprising two nuclei, is deceptively simple. But the magic of quantum mechanics turns even this object into a highly interesting affair. The pair of spins when placed in an external field can exist in four states: up-up, down-down, up-down and down-up or their linear combinations. Similar to the case of a single spin, these four states have different energies.

There is another simple but far-reaching rule operative in the quantum statistical world: the ‘Pauli principle’. The principle explains the formation of black holes and neutron stars and also describes when and how hydrogen atoms can come close together and recombine in making something as simple as a hydrogen molecule.

In simple words, the Pauli principle states that particles can recombine only when the resulting molecule obeys certain symmetry considerations. The molecule is not just the spin of the nuclei; it also comprises the spins of the electrons and the translation, vibration and rotation of the molecule as a whole. Each of these properties is called a “degree of freedom”. Every degree of freedom possesses its own symmetry and in the molecule, the components must combine to impart the whole molecule with a particular symmetry.

For example, supposing we have two variables x and y and would like their product to be negative, that is, to possess odd symmetry. Clearly, an odd (or negative) x must combine with an even (or positive) y and an even x must combine with an odd y. Like combinations such as both variables being odd and both being even are not permitted. Conversely, if y is even, we can always be sure that x must be negative. In other words, we have forced x to take up a certain state dependent on y. The degrees of freedom become correlated.

Likewise, in the hydrogen molecule, we can force all the nuclei to exist in just one state if we can choose the other degrees of freedom. The energy gaps in the nuclear spins are very small and extremely difficult to manipulate, but the gaps in the molecular motion are very large and it is easy to redistribute molecules in one level or the other. This fact helps us choose the correct molecular degree of freedom and the Pauli principle ensures that all nuclear spins are in the same state.

With “hyper-polarisation”, we have now entered a new paradigm of MRI. Hydrogen nuclei can now be polarised with a fraction of one in one, rather than one in a million, about a million-fold improvement. The hydrogen prepared in this special way, called “para-hydrogen”, can arguably pave the way for a radically new outlook in MRI and spectroscopy. In MRI, we can acquire higher contrast, sharper images and in spectroscopy, we can now shed light on the dark areas and see the otherwise hidden trace metabolites.

The lung tissue, for example, has very poor hydrogen content. In an MRI scan a piece of lung will be almost invisible. However, air mixed with para-hydrogen can be inhaled and the tissue simply “glows up” enabling detailed scans. It is also possible to measure the rate at which the gas moves through blood and how different tissues in the brain take up this gas, or possibly how heart tissue behaves in congestive heart failure.

Also, we are not limited to hydrogen. The hyper-polarisation can also be transferred to other nuclei such as carbon and phosphorus, enabling highly revealing and sensitive non-hydrogen MRI scans. The options are quite limitless enabling us to ask questions we have never been bold enough to ask before.

For example, could this as yet unexplored avenue in hyperpolarised functional magnetic resonance open up new vistas for investigating the very metabolic foundations for human thought and consciousness? Could this illuminate the biochemical basis of disease and suggest intelligent therapies? Can our perfect polarisations even absolve us from the need to use expensive and bulky superconducting magnets, bringing MRI within the reach of the common man.

The ‘silent’ spins are silent no more. So listen to their symphony, for after all, hearing is believing!

The writer is a freelance contributor

Please Visit our Sponsor (Ads open in separate window)