GENERATION AND DETECTION OF NMR SIGNALS
The simplest NMR spectroscopy or imaging system consists of a magnet with field gradients, a transmitter, a receiver, a coil/probe, and a computer that collects the data and orchestrates the timing and use of the entire system. The following sections explain the interaction of these components and discuss their functions and engineering equipments for high-quality clinical applications of NMR.
The performance and quality of NMR spectrometer or imager dependent on the magnitude and homogeneity of the static or direct-current (dc) magnetic field. Initially most magnets used in the development of NMR were either permanent or electromagnetic (resistive). As higher magnetic field strengths were required, super-conducting magnets appeared and are now the industry standard. Resistive magnets have become less important not only due to their lower field strengths but because they require excessive amounts of electrical current to maintain their fields.
Superconductivity is the interesting property of certain materials that when cooled to liquid helium temperatures become perfect conductors. By a perfect conductor we mean that electrical resistance in the wire not just small, but is absolutely zero. Initially production of high magnetic field with this technology was not only limited by the critical temperature of their super-conductivity, but also by a unique critical magnetic filed magnitude. These field magnitudes were very low. Fortunately a new class of alloys of the metal niobium was discovered to be able to carry very large current densities. This increase in current-carrying capacity resulted in today’s high-field spectroscopy and imaging systems. Curiously it is the relation among critical temperature, filed and current density that is obstacle to using recently discovered liquid-nitrogen-temperature superconductors for high-field magnets.
Less than 3% of the more than 2000 magnets operating in the United States are low-field systems. These systems operate at fields far below super-conducting magnets having field strengths of 0.064 – 0.3 tesla. Uniform magnetic fields can be achieved at the center of permanent or electromagnet, if the width of the gap is much smaller than the lateral dimensions of the magnet.
Room temperature magnetic-gradients are used in two ways in NMR imaging. They are used either to compensate for magnetic-filed inhomogeneities of the magnet or to provide an encoding magnetic-gradient to make an image. In either case these gradients are arrangements of loops of copper wire designed to produce well-defined gradient patterns in the bore of the magnet.
Perturbation of the magnetisation vector is accomplished with an external RF energy source at the same frequency as the Larmor frequency of the nuclei. The RF energy is coupled to the object containing these nuclei by a RF inductor (NMR coil). Since NMR devices are typically operated in pulsed mode, the same coil may be used following excitation for detecting the NMR signal emitted by the nuclei returning to the lower energy state.
Surface coils are simple wire loops of various sizes tuned and matched to the frequency of interest. These coils overcome two technical difficulties in NMR spectroscopy and imaging. First, surface coils are useful because they localise signal reception to organs or structures of interest. Second, they provide greater sensitivity since their radio-frequency magnetic field is more intense than larger homogeneous coils.
All pulse NMR applications demand a short but intense high-power RF magnetic field in order to perturb the alignment of the nuclear magnetisation to obtain a signal. A frequency generator provides a source of continuous RF input. This signal is gated to allow the computer to control the timing of the pulse at the proper 5-20 KW of output power. This amount of power could be very damaging if the power remained switched on for any reason.
The receiver consists of a number of individual components. The incoming NMR signal has already been amplified by the preamplifier as it emerges from the coil. In receiver the signal is first demodulated. This means that the earlier frequency, which is either near or equal to the resonance, is removed.
The two analog signals are filtered once more to eliminate aliasing and are digitally sampled. The data is sampled (on the order of 50 KHZ) so that there are at least two data points for the highest frequency cycle in the sample. This referred to as the Nyquist condition. If this condition is ignored frequencies may appear in the Fourier transformed data set at one sampling rate below the true frequencies.
