Esemplastic thermoplastic

Back in 1946, the physicists Bloch and Purcell tried to build a device that would let them directly measure magnetic moment of a nucleus. It was based on a simple idea: the energy level of a nucleus placed in a strong magnetic field splits by exactly μB₀, where B₀ is the applied magnetic field, and once it is split, they would scan the RF frequencies until the find one that matches the gap. Infortunately, their device turned out to be useless; they didn’t realize that those pesky electrons around each nucleus shield it from B₀ and the actual field that splits the energy levels is always less than that, less by some unknown number.

Little did they know what was an unknown number for a physicist would be a revolution in chemistry and would bring them a Nobel Prize in 1952! That number was the chemical shift, which became the x-axis of every Nuclear Magnetic Resonance spectrum, the ubiquitious analysis used in every field of chemistry and biochemistry. The NMR spectra let the chemists identify every single atom in a molecule, tell them which atoms it is connected to, tell them how far are pairs of atoms from each other in space (thus reporting 3D coordinates), and provide additional information on dynamic behavior and intermolecular interactions.

One thing that has annoyed every chemist who had to do NMR is that the sample must be perfectly homogeneous, and the magnetic field must be absolutely uniform throughout the sample, undisturbed by impurities, vessel walls, and such. Every NMR operator spent the majority of his working time shimming the field, which, in NMR jargon, means making small ajustments to the correction coils to make the field in the sample perfectly uniform. This annoying problem for the chemists turned out to be nothing less than another Nobel Prize winning revolution, now in medicine. Lauterbur and Mansfield in 1973 figured out how to use nonuniform magnetic fields to record the spectra independently from every given point within the sample, and that made it possible, by taking an NMR spectrum of ordinary water, to look inside a living human body using non-ionizing radiation.

Magnetic Resonance Imaging, or MRI, evolved into many amazing applications, allowing to see more than just tumors and tissues – there is now FMRI which lets us directly see things such as neural activity in living, conscious human brain, there is MRI angiography, there is Diffusion-Weighted Imaging to study post-stroke cytotoxic edemas. MRI can be used to guide surgery and to view extremely small details in Magnetic Resonance Microscopy. It is truly a great scientific advance, and it’s only even more amazing that it grew out of the scientists questioning their failures.

Let’s recap the sciences working for you each time you get your MRI done:

• Physical sciences:
  • Quantum Mechanics: The NMR itself is a quantum mechanical experiment, where the solution to the Schrödinger equation for Zeeman Hamiltonian is calculated, and where quantum states of the nuclei evolve and change when they become subject to the quantum operators introduced by the pulse sequences.
  • Physics of superconductors: A practical challenge with NMR is building a magnet providing fields of 10, 20, or even, in the most recent experimental setups, 30 Tesla. Such field density is only possible with a superconducting solenoid, which brings along the complex quantum physics of the superconducting state with its Cooper pairs and Meissner effect and the like.
  • Electromagnetistm: not only the humongous magnetic field must be energized, stabilized, and shimmed to uniformity, NMR spectrometers have radio frequency transmitters and detectors, detectors so fine they can pick up the signals down to their natural Heizenberg uncertainty, signals many orders of magnitude smaller than the thermal energy of the sample. They often have to be cryocooled themselves, to suppress line noise.
• Chemical sciences
  • NMR is a tool of chemistry, yes a lot of unusual chemistry goes into its creation. The commonly used superconductors such as Nb₃Sn or NbTi bring along the challenges of manufacture and efficient use. NMR has already approached their limits and high temperature superconductors, such as Bi₂Sr₂CaCu₂O, are now sometimes used for innermost parts of magnets over 25 Tesla, due to their high critical current density even at such fields.
• Computer sciences
  • NMR is impossible without extensive computational power. The signal received from the RF detector must be weighted, fourier-transformed (possibly many times, for multidimensional spectra), post-fourier filters may have to be applied, before a sensible spectrum can be seen. After that, if working with complex organic or biochemical molecule, massive amount of work goes into assignment of the spectra and even more into the calculations of protein structures or similar complexity goals.

Incidentally, NMR spectroscopy (not MRI) was the main focus of all my college work, my Ph.D. thesis and postdoctoral research. I think that MRI, the ability to see inside a living human being, is one of the greatest advances of modern science.