The MR Scanner
Magnetic resonance imaging builds detailed pictures of the body's interior using nothing more than strong magnetic fields and radio waves — no ionising radiation anywhere in the process. It works because every hydrogen nucleus in the body is, in effect, a tiny compass needle that can be made to broadcast its own location.
An entirely different route to an internal image
Magnetic resonance (MR) imaging produces detailed images of internal body structures using strong magnetic fields and radio waves, and involves no ionising radiation whatsoever — a genuinely distinct physical principle from every other imaging technique covered in this module. Where ultrasound relies on mechanical pulse-echo reflection and other methods rely on transmitting or emitting ionising radiation through the body, MR imaging instead exploits a magnetic property of the nuclei that make up the body's own tissue. This lesson covers the core physical mechanism behind MR imaging — proton alignment and precession, gradient fields, and RF signal detection — at the level actually needed to understand how a scan is formed. The detailed physics of relaxation times, sometimes covered in more depth elsewhere, is not needed here.
Proton spin alignment and precession
Hydrogen nuclei — single protons — are extremely abundant throughout the human body, largely because the body is mostly water and fat, both rich in hydrogen. Each proton possesses an intrinsic property called spin, and as a result behaves like an extremely small magnet, with its own magnetic axis.
Normally, the magnetic axes of all these countless protons point in random directions throughout the body, and their tiny magnetic effects cancel out. Once a patient is placed inside the scanner's very strong, highly uniform magnetic field, produced by a superconducting magnet, this changes: the protons' spins tend to align with the applied field — a small majority aligning with the field direction rather than against it, since aligning with the field is very slightly lower in energy. At the same time, rather than sitting perfectly still, each aligned proton precesses — wobbles, in much the same way a spinning top wobbles around the vertical while gravity acts on it — around the direction of the magnetic field. The frequency of this precession depends directly on the strength of the magnetic field the proton actually experiences at its location.
Gradient coils and spatial encoding
If every proton in the body experienced exactly the same field strength, every proton would precess at exactly the same frequency, and there would be no way to tell which part of the body a detected signal had come from. This is solved using gradient coils: additional coils inside the scanner, separate from the main superconducting magnet, capable of superimposing a small, deliberately varying (gradient) magnetic field on top of the main uniform field.
With a gradient field switched on, the total magnetic field strength varies slightly and predictably with position along the body — and because precession frequency depends directly on local field strength, the precession frequency also now varies slightly and predictably with position. This position-dependent precession frequency is the key that allows spatial information to be recovered later: a detected signal at a particular frequency can be traced back to a particular position in the body, because the gradient field made that frequency unique to that location.
RF excitation and de-excitation signals
To actually generate a detectable signal, a radio-frequency (RF) pulse is applied, tuned to match the precession frequency corresponding to a particular field strength — and therefore, given the gradient field in use, to a particular location in the body. Protons precessing at that specific frequency absorb energy from the RF pulse, which disturbs their alignment with the main field.
Once the RF pulse switches off, those disturbed protons realign themselves back with the main magnetic field. In doing so, they release the energy they had absorbed, re-emitting it as a detectable RF signal of their own — a de-excitation signal. It is this emitted signal, not the original applied pulse, that the scanner's receiver actually detects and measures.
Computer reconstruction
Because the frequency of each emitted signal depends on the position it came from (via the gradient fields), signals detected from many different positions, angles, and gradient settings across the body can be collected and combined. A computer processes this large collection of frequency-encoded signals, working out where in the body each contribution originated, and reconstructs a full two-dimensional or three-dimensional image of the internal structure — all without ever needing to physically look inside the patient or pass any ionising radiation through them.
MR imaging shares a central advantage with ultrasound — no ionising radiation is involved at all — but achieves it in a completely different way. Where ultrasound avoids ionising radiation because it isn't a form of radiation to begin with, MR imaging avoids it because its signal comes entirely from magnetic fields and radio waves, which do not carry enough energy to ionise atoms in the body. This makes MR imaging comparatively safe for repeated scanning, without the dose-accumulation concerns that come with X-ray or radionuclide-based techniques.
Key takeaways — Ultrasound, Endoscopy and MRI
- Ultrasound generation and detection both rely on the piezoelectric effect: the converse effect turns an applied voltage into vibration (generating a pulse), and the direct effect turns a returning vibration into a voltage (detecting an echo).
- Acoustic impedance, $Z = \rho c$, governs how much of an ultrasound pulse reflects at a tissue boundary, via $\dfrac{I_r}{I_i} = \left(\dfrac{Z_2-Z_1}{Z_2+Z_1}\right)^2$ — the physics behind every ultrasound echo.
- A-scans convert echo timing into a single depth measurement; B-scans combine many such lines, coded by brightness, into a full 2D image — set against other modalities on safety, resolution, and cost.
- Fibre optics uses total internal reflection at the core-cladding boundary to guide light, and coherent bundles (unlike non-coherent ones) preserve a spatial image, enabling direct visual endoscopy.
- MR imaging aligns and precesses hydrogen protons in a strong magnetic field, uses gradient coils to make precession frequency position-dependent, and detects RF signals re-emitted as protons realign after an RF pulse.
- Computer reconstruction turns these position-encoded RF signals into a full image, entirely without ionising radiation.
- Together, these five lessons form a complete tour of non-X-ray, non-radionuclide diagnostic imaging physics — mechanical (ultrasound), optical (endoscopy), and magnetic/radio-wave (MRI) routes to seeing inside the body.