The Hydrogen Nucleus: Water and Fat

The ¹H Nucleus: Water and Fat

The majority of MRI techniques measure the tiny signal arising from transitions between magnetic energy levels (associated with different ‘spin states’) of the hydrogen nuclei in water. Because the ¹H hydrogen nucleus comprises just a single proton ¹H MR techniques are often referred to as ‘proton MR’ even though all atomic nuclei contain protons. Proton MR does not measure a signal from any protons other than those in hydrogen nuclei. To perform any kind of imaging it is necessary to know the spatial origin of the detected signal. In MRI the spatial position of a hydrogen nucleus in the imaged object is ‘labelled’ with a particular resonance frequency and phase by imposing well-defined short-duration magnetic field gradient pulses in addition to the main field. Even at the highest achievable spatial resolution the measured MR signal for each voxel is the average signal from a collection of billions of nuclei.

Two factors make the MRI signal very weak. Firstly, the energy involved in a transition between nuclear ¹H magnetic energy (‘spin’) states is ten orders of magnitude less than that of an X-ray photon. Secondly, detectable transitions arise only due to the very slight excess of nuclei in the lower energy state. This excess depends on both the temperature and the magnetic field strength and is only ≈ 0:0005% at body temperature in a 1.5 Tesla MRI magnet. Fortunately, there are plenty of hydrogen nuclei available in biological tissue: ≈5 x 10²² in a cubic centimetre. The energy difference between spin states is proportional to the magnetic field strength. These two factors, population and energy difference, are the reason for the gradual replacement of 1.5 T MRI scanners by 3 T and higher field systems – they produce more signal.

In the absence of the MRI magnet, only the Earth’s magnetic field (≈0.00005 T) would be present. Both the population and energy differences between the spin states would be very much smaller and no clinically useful signal would be available. Nevertheless, using a small ‘prepolarizing’ electromagnet to boost the population difference, it is possible to perform low-resolution MRI in the Earth’s field.

In biological tissue, the next most common hydrogen nuclei after water are those present in fat or lipid, especially the -CH₂- groups that form the backbone of fatty acids. In general, the fat signal does not contain any diagnostically useful information, and because the hydrogen nuclei in lipids resonate at a slightly different frequency from those in water, the part of an MR image arising from a lipid may be displaced relative to the adjacent tissue water image. This is a chemical shift artefact. It can be avoided by suppression of the lipid signal during the MRI measurements.

There are many different ways of generating contrast in MRI, each depending on a particular physical property, or a mixture of physical properties, of tissue. All MRI techniques depend on the input of RF energy, manipulation and evolution of the stimulated system, and measurement of the RF energy emitted as the system relaxes back to thermal equilibrium. In MRI the term ‘thermal equilibrium’ refers to the populations of nuclear spin states that exist at a particular temperature and magnetic field strength. Absorption of RF energy disturbs this equilibrium. The emission of RF and other energy transfer processes restore equilibrium. Disturbance and restoration of thermal equilibrium, at least in MR measurements, does not mean the temperature of the sample is going up and down significantly. However, the small proportion of absorbed RF energy that is converted to heat can cause significant tissue heating in RF-intensive MRI pulse sequences.

The simplest form of MRI contrast is dependent on the amount of water present in a particular tissue or, strictly speaking, the water hydrogen nucleus density. These are called proton density (PD) images. Since the range of proton densities in tissue varies only in the range of 70–100% compared with pure water not much contrast can be generated. PD images are, however, particularly useful for the examination of cartilage.

The rate at which the measurable signal from this collection of nuclei decays depends on a large number of factors in the local molecular environment of each nucleus. If the ‘system’ (the collection of ¹H nuclei) is not manipulated in some way, by the application of extra magnetic fields or extra RF energy, then the measurable RF signal will generally decay to zero within one second of input of the first pulse of RF energy. Differences in this rate of signal decay (T2 relaxation) provide one method of MR contrast generation (T2 contrast).

The lack of a measurable RF signal does not mean that a stimulated system has relaxed back to thermal equilibrium. It can simply mean that the emitted signals are incoherent – they cancel each other out. To a limited extent, the system can be manipulated to recover some coherence but ultimately no RF signal can be detected because all the input energy has been reemitted or converted to heat. Thermal equilibrium of the system is achieved (T1 relaxation) at a rate that is also dependent on the local molecular environment, though due to factors not identical to those that cause T2 relaxation and signal incoherence. Differences in the rate of equilibration of local environments can also be used as a contrast mechanism (T1 contrast). The time required for water protons in biological material to reach thermal equilibrium is typically 5–10 s

Because the position of nuclei in MRI is labelled with their frequency and phase, nuclei that change position in the imaged object between the time of labelling and the time of measurement may give a ‘spurious’ or ‘incorrect’ signal. Such molecular movements might be due to fluid flow, simple diffusion, or patient movement. In the case of patient movement, the result is often a distinctive movement artefact evident in the image. Flow and diffusion, on the other hand, can cause a signal increase or decrease depending on whether signal coherence is lost or gained. Diffusion-weighted and perfusion-weighted imaging methods develop contrast according to the physical mobility of water in tissue and body fluids.

The structural components of cells and tissues affect the microscopic and macroscopic molecular environment and thus the MR signal from water indirectly reflects both body structure and function. If the tissues of interest in a particular examination do not have good inherent contrast then the local molecular environment can sometimes be modified by the administration of MR contrast agents. These are mostly based on paramagnetic metals that affect the T1 and T2 relaxation rates.