MR Image Encoding

Much of the success and flexibility of MRI is derived by its peculiar methodology; a set of techniques which has proved to be very flexible and informative for probing the properties of complex materials (such as the brain). MR imaging is fundamentally different from other types of imaging. Conventionally, when physicists refer to “imaging” they refer to a scattering experiment. Since our eyes “see” a rock by forming a 2 dimensional array of light intensity amplitudes being scattered off the rock, the instrumental equivalent of this has come to be synonymous with imaging. In the canonical scattering experiment, “rays” are aimed at an object and then detected as they either scatter off of the object or penetrate through it. The rays can be deflected, lose or gain energy, deflect and then interfere with one another or even be converted from one type of ray to another. The “rays” in question are usually electromagnetic radiation (radio waves, microwaves infra red light, visible light, ultraviolet light, x-rays, or gamma rays) but can, of course, be almost anything including sound waves, water waves, or matter waves (particles). The particles used could be electrons (as in electron microscopy), or any of the zillions of other subatomic particles, or clumps of particles such as nuclei, atoms, molecules, or pieces of dirt. A fundamental limitation of scattering experiments is that the resolution of the image cannot exceed the wavelength of the wave (electromagnetic or matter wave) used to probe the object. This is fundamentally derived from the uncertainty principle but has long been understood in classical optics. Physicists are fond of making such general statements, but, while true for scattering experiments, this imaging “law” does not hold for MRI. In MR, we use radio waves with a wavelength of several meters to image at a sub-millimeter resolution. Thus we exceed this “fundamental” imaging law by several orders of magnitude. How? The short answer is that we don’t determine the distribution of the body’s protons by bouncing stuff off of them; we determine it by asking them to report to us where they are. We query the body’s spins using a burst of radio waves (the excitation RF), and they report back some milliseconds later with a faint radio signal of their own. They declare their location and other even more valuable information encoded in the frequency and phase of the burst of RF energy they emit in response (the MR signal or echo).