Starting points

The invention of the principle of imaging using nuclear magnetic resonance (NMR) signals by Paul C. Lauterbur at the State University of New York in Stony Brook/Long Island in 1973 sparked a radical change in the world of medical imaging, offering as it did the possibility to capture cross-sectional images of the body without exposing it to harmful radiation. In the early 1980s, however, when commercial manufacturers of medical equipment began to produce the first MRI scanners, the examination was a stressful experience for patients, who had to remain motionless in a narrow tube for several minutes at a time. This was generally felt to be an insurmountable obstacle, the expectation being that MRI scanning would never be as fast as X-ray computed tomography (CT).

The potential of MRI was, however, recognised at the world-famous Max Planck Institute for Biophysical Chemistry in Göttingen, where a biomedical NMR group headed by Jens Frahm was set up in 1982 to develop scientific and medical applications of the technology. Frahm and his team focussed on the problem of reducing data acquisition times, since this was the main factor keeping MRI from being used on a large scale. After years of research, Frahm’s team discovered a ground-breaking new technology for speeding up the process by at least two orders of magnitude. With Fast Low Angle Shot (FLASH) MRI, the acquisition time for individual images was reduced from several minutes to a few seconds.

Technical Background

What is MRI?
Magnetic resonance imaging is based on the fact that the human body is largely composed of water molecules, each containing two hydrogen atoms. Their charged nuclei, called protons, have a spin which generates a magnetic moment.

Figure 1

Fig. 1: Spin and magnetic moment of a proton*

Inside an MRI scanner, the protons are exposed to a strong and uniform magnetic field which aligns their magnetic moments with the direction of this field. The ensemble of all magnetic moments leads to a macroscopic magnetisation which marks the equilibrium state of the proton spins.

Figure2

Fig. 2: Equilibrium state of proton spins*

The application of a radiofrequency excitation pulse then flips a number of proton spins which – from the perspective of classical physics – results in a rotation of the macroscopic magnetisation by a pre-determined angle (“flip angle”) away from the direction of the field. The flip angle depends on the amplitude and duration of the pulse. The excited magnetisation precesses around the direction of the static magnetic field and induces a voltage (“NMR signal”) in the receiver coils of the MRI scanner.

This time-domain signal is transformed into frequency space, where it leads to a single frequency representative of the water protons in the body and the strength of the applied magnetic field. Paul Lauterbur discovered that this frequency dependence of the signal can be used to generate an image if a magnetic field gradient is applied during data acquisition. Because the frequency linearly depends on field strength, a signal recorded in the presence of a magnetic field which varies along one spatial dimension automatically results in a frequency distribution which emerges as a one-dimensional signal projection of the object under investigation. In order to measure enough data for an image, such spatially encoded experiments have to be repeated many times, with different magnetic field gradients. This need for multiple repetitive experiments is the first problem determining the long MRI acquisition times.

Figure3

Fig. 3: Rotation of macroscopic magnetisation using an excitation pulse*

The second problem is due to the fact that, after each excitation, the excited magnetisation has to return to its equilibrium state. This relaxation process is characterised by two tissue-dependent relaxation times which are responsible for the contrast in the resulting image. Most importantly, however, the proton relaxation time which determines the exponential recovery of the equilibrium magnetisation is of the order of one second. It therefore causes a long waiting period of several seconds before a subsequent experiment with a different spatial encoding can be performed with sufficient signal strength. For example, 256 repetitive experiments with a two-second repetition time result in a measuring time of 512 seconds - almost nine minutes - for one image.

Problems prior to the invention of FLASH
In the conventional MRI technology in use prior to FLASH, the flip angle for excitation was set at 90 degrees in order to maximise the MRI signal picked up by the receiver coils of the MRI scanner. However, as this flip angle excites the entire available equilibrium magnetisation, the delay between two consecutive experiments has to be long enough to allow the macroscopic magnetisation to return to its equilibrium state. This delay is in the order of several seconds. Otherwise a saturation effect will occur such that no useful signal can be acquired anymore during repetitive excitations. Hence, conventional MRI technology prior to FLASH was unsuitable for rapid data acquisition.

It should also be noted that the most promising conventional MRI methods used a second pulse with a flip angle of 180 degrees. This pair of pulses rephased the excited magnetisation in a so-called spin echo signal, which was generally accepted as advantageous for MRI as it was not affected by inhomogeneities of the static magnetic field. Spin-echo images were therefore free from artefacts such as local geometric distortions or signal losses.

The advantages of FLASH
Frahm and his team achieved a huge acceleration in image acquisition times using excitation pulses with flip angles of much less than 90 degrees, in most cases between 5 and 15 degrees. Accordingly, the new method was called “Fast Low Angle Shot” (FLASH) MRI. In order to fully benefit from this trick, Frahm’s team also had to abandon the conventional MRI technology. This meant eliminating the established spin-echo technology, because acceleration by low flip angle excitation only works if no following pulse further manipulates the proton spins. The FLASH technology therefore acquires the MRI signal in the form of a so-called gradient echo which may be generated by reversal of a magnetic field gradient after a single low flip angle pulse.


* Images based on: Weishaupt, D., Köchli, V. D., Marincek, B. (2014): Wie funktioniert MRI? Eine Einführung in Physik und Funktionsweise der Magnetresonanzbildgebung. Berlin Heidelberg: Springer.

Last modified: Tuesday, 6 March 2018, 3:14 PM