MRI creates an image by detecting radiofrequency signals from atomic nuclei, usually hydrogen in the body’s water and fat. A strong magnetic field establishes the conditions for those signals; radiofrequency pulses disturb the magnetisation; changing magnetic fields encode position. A computer reconstructs the measurements into an image.
The important missing step in many explanations is position encoding. Detecting a radio signal alone would not tell the scanner which part of the body produced it.
Four jobs that must happen together
1. Establish a small collective imbalance
Hydrogen nuclei behave magnetically. In the scanner’s main field, a small imbalance between their possible energy states produces a net magnetisation. Saying that “all the protons line up” is a useful first picture but an inaccurate literal description: the measurable effect comes from a collective imbalance, not every nucleus pointing identically.
The NIH explanation of MRI describes how the main field and radiofrequency excitation make tissue signals measurable.
2. Excite and receive
A radiofrequency pulse changes the orientation of the net magnetisation. After excitation, its changing transverse component induces a voltage in receiving coils. The equipment measures this signal while the magnetisation evolves. It does not take a tiny optical photograph of each proton.
3. Give location a magnetic signature
Gradient coils make the magnetic field vary with position. Because the nuclei’s precession frequency depends on field strength, gradients allow frequency and phase to carry spatial information. Slice selection, phase encoding and frequency encoding use this relationship in coordinated ways.
NIST’s account of MRI measurement helps connect the magnetic behaviour to the scanner’s need to distinguish locations. The acquired data are commonly organised in a mathematical space called k-space. One measurement is not simply one finished image pixel.
4. Reconstruct and choose contrast
The reconstruction combines many measurements to estimate the spatial distribution of signal. Pulse timing and sequence design determine which tissue differences become prominent. Longitudinal recovery, transverse signal decay and hydrogen density all matter; image brightness is not a universal tissue label.
| Misreading | More useful interpretation |
|---|---|
| White always means the same material. | Brightness depends on the sequence and display settings. |
| The scanner records each pixel directly. | Spatial information is encoded into measurements and reconstructed. |
| A stronger magnet removes all image limitations. | Motion, sampling, coils, sequence design and artefacts still matter. |
Why motion matters even when the scanner sees inside you
Measurements are collected over time. Movement changes the relationship between the encoded positions and the anatomy, so combining them can produce blur or ghosts. This is a different timing problem from a rolling shutter distorting a moving object, but both show why an image is not always a single instant.
MRI uses non-ionising radiofrequency energy rather than the X-rays used in CT and cone-beam CT. That distinction does not make every MRI situation risk-free. The FDA’s MRI information explains the importance of magnetic fields, implants, heating and screening. Suitability and protocol choices belong with the imaging team; the physics alone cannot select a scan for an individual patient.

Hi, this is a comment.
To get started with moderating, editing, and deleting comments, please visit the Comments screen in the dashboard.
Commenter avatars come from Gravatar.