MRI—magnetic resonance imaging—is a powerful medical technology that has revolutionized diagnosis of a very wide range of illnesses and injuries, greatly reducing or in many cases eliminating the need for exploratory surgery. It provides medical practitioners with two- and three-dimensional images, as well as high-accuracy cross-section, of internal structures and organs within a patient’s body.

Underpinning the results achieved by MRI scanning is a wide range of advanced technologies, including precision measurement techniques: the almost unbelievable sharpness of the pictures that MRI produces depends directly on measurements of basic electrical parameters.

Figure 1: Atomic nuclei of atoms have a magnetic moment, represented a vector quantity with its direction along the rotation axis.MRI frequently sits alongside, and in some ways is complementary to, computer tomography (CT). CT scans are based on X-rays and are best at imaging high-density structures (such as bones), whereas MRI scanning reveals the details of soft-tissue structures.

The working principle of MRI is based on nuclear magnetic resonance (NMR). In fact, what MRI actually detects is the magnetic resonance of the protons of hydrogen atoms contained in the water within the human body: water represents up to 70% of body weight. In more exact terms, MRI observes the response of the hydrogen nuclei exposed to excitation by both magnetic and electromagnetic fields.

The collected energy per volume element (voxel) depends on the water distribution in the place under analysis. So MRI can provide a three-dimensional image of the water distribution inside the human body. As each type of body tissue has a characteristic proportion of water within it, it becomes possible to image those tissues, and any deterioration, by looking at changes in water distribution.

Working Principle of the NMR
The nuclei of atoms have the property of behaving like magnetic dipoles or magnets when excited by a magnetic field (Figure 1). Nuclei of atoms have a spin (or magnetic moment) which we conventionally represent by a vector along the rotation axis.

Figure 2: With a DC magnetic field H0 between 0.2 and 3 [T], the spins are in line with the field.In the absence of any external influence, this tiny magnet is not oriented in any particular direction. As soon as this magnet is illuminated by a constant and homogeneous static magnetic field (referred to as Ho) it aligns with Ho in two directions: parallel and anti-parallel to the field. The nuclear magnetic moment is tiny and requires an intense applied field to achieve the alignment; the related magnetic induction Bo is commonly between 0.2 and 3 Tesla. In Figure 2 (presented as a necessarily simplified explanation), only the parallel alignment is considered.

Figure 3: At any instant, the spin axis is not aligned with the applied field but due to its precession the average x and y components cancel out.The alignment process is more subtle than a simple setting of the spin axis along the field lines. If we take the z-axis (Figure 3) as parallel to the applied field, the spin precesses or rotates around the z axis along a cone at angular speed w0. The related frequency is called the Larmour frequency: w0=gB0.

The precession speed is therefore proportional to the static magnetic field; for example, a field of Bo=1Tesla gives a frequency f0 = 42.5 MHz.

Resonance of the Nuclei
In order to observe the resonance of the nuclei, some energy has to be provided allowing nuclei to move from steady state to excited one. This is achieved by applying a high frequency magnetic field H1. When the frequency of H1 equals the Figure 4: An excitation antenna excites the nuclei with a frequency matched to the Larmour frequency.Larmor frequency, resonance occurs and the nuclei move to a higher energy state.

During the application of H1, the spin axes of the nuclei are no longer aligned with Ho (z axis) but move into the x-y plane. After the H1 excitation is turned off, the spin axes once again align with H0, and the extra energy they gained from the H1 excitation radiates away in the form of a damped electromagnetic wave (also known as relaxation). An antenna detects damped wave, yielding an induced voltage called Free Induction Decay (FID).

It is the FID signal that the MRI’s computer processes to a 3D or section image (Figure 5)

Applying the Magnetic Fields
Figure 5: Atomic nuclei radiate energy as they realign with the static magnetic field, allowing their distribution to be mapped.The Static magnetic field Ho, as previously noted, must be very intense, with very high stability and homogeneity within the volume inside the aperture of the MRI scanner, where the patient lies.

Most of today’s MRIs generate the static field by means of a superconducting magnet located around the cylinder of the scanner. The coils of the magnet are made up of niobium-titanium (NbTi) wires immersed in liquid Helium at a temperature of 4K.

The gradient coils superimpose a magnetic gradient to Ho in order to provide a spatial coding of the image. Imaging takes place in just one plane or slice at a time, and to ensure that signals are received only from nuclei in that plane, only those nuclei have to be pushed to resonance.

Figure 6: Gradient coils add to the static field at one end and diminish it at the other, controlling the plane in which the total field has exactly the correct value.The appearance of the resonance is strongly dependent on the value of the magnetic field Ho: the gradient coils superimpose a magnetic field to ensure that the final magnetic field is exactly equal to Ho only in the plane of interest.

How the Gradient Coils Work
To create a gradient along an axis, a pair of coils is needed. In each pair, currents flow in opposite directions (Figure 6).

In fact, three pairs of gradient coils are located around the cylinder of the MRI apparatus to create three orthogonal magnetic fields. So, it is possible to adjust the magnetic field at any point in the volume of the cylinder. Gradient amplifiers operating in a closed servo-type loop drive the currents in the gradient coils (Figure 7). Each MRI therefore needs three such current control loops.

Figure 7: Feedback from output current transducers is fundamental to obtaining the required degree of precision from the gradient current amplifier.As can be seen from the principle of MRI outlined in Figure 7, the quality, the clarity, and the resolution of the images are directly linked to those of the magnetic field applied, and therefore to those of the current injected into the gradient coils. One of the key elements in the current control loop is the global accuracy of the current transducer.

In particular, the following parameters of the current transducer are critical:

  • Extremely low non linearity error (<3 ppm of measuring range)
  • Very low random noise (low frequency noise from 0.1 to 1.0 kHz)
  • Very low offset and sensitivity drifts over temperature range (<0.3 ppm/K)
  • Very high stability of offset versus time (one reason for this is the duration of MRI scans, some of which may last several tens of minutes)
  • Measuring range (around 1000 A peak)
  • Bandwidth (–3 dB point of 200 kHz)

To reach these performance levels, Hall Effect current transducers, which were used in previous generations of MRI scanners, are no longer adequate. The solution developed by LEM, primarily for this application area, has similarities to the Hall Effect technique but offers significant advantages. It is described as a double fluxgate closed loop transducer and identified as type ITL 900. Although fluxgate technology has been available for some time, LEM was able to take this technology and adapt and improve it.

As well as precise current control in gradient amplifiers for medical imaging, the ITL 900 is equally applicable to measuring feedback in precision current regulated power supplies, current measurement for power analysis, calibration equipment for test benches, and laboratory and metrology equipment which also require high accuracy.

In its present form, the technology is limited to a relatively narrow operating temperature (typically +10°C to +50°C). However, LEM is confident that the technology can be developed further and that the ITL 900 transducer could prove to be as significant for the future of MRI scanning as the Hall Effect transducers were for its introduction. As with Hall Effect itself, with its leading edge performance ITL900 could possibly enable any number of future applications.