Nib asked, and Nib — askNib's tutor — drew the answer live at a whiteboard. This is the spoken transcript; enable JavaScript to watch it drawn.
how does an mri machine work
An M R I scanner takes your picture with no X-rays at all. It uses a huge magnet and radio waves to listen to the hydrogen in your body's water and fat.
The magnet is the heart of it. Most scanners run at one and a half tesla, tens of thousands of times Earth's own field, held by superconducting coils kept cold with liquid helium.
Every hydrogen nucleus is a single proton, a tiny spinning magnet. In the big field it wobbles like a tilted top, at 42.58 megahertz per tesla: about 63.9 million times a second here.
Now a radio pulse at exactly that frequency tips them over. Only matching radio waves get absorbed. When the pulse stops, the tipped protons hum back, and a coil hears it.
After the pulse, each tissue regrows its magnetism at its own pace, called T1. At one and a half tesla fat takes about 260 milliseconds, brain white matter 780, spinal fluid 2,500.
Pulse again after half a second. Fat has regrown most of its magnetism, fluid barely any. So fat shines and fluid looks dark, and that contrast is the picture.
But where is each signal from? Gradient coils tilt the field so it grows steadily along the body. Stronger field means faster wobble, so frequency now tells position.
The coil hears every frequency at once. A computer unmixes them with a Fourier transform, and the loudness at each frequency becomes the brightness at each place. The knocking you hear is the gradient coils flexing.
So the whole machine in one picture: the magnet lines protons up, a radio pulse tips them, gradients tag each place, the coil listens, a computer builds your image.
An M R I is a conversation with hydrogen: tip it with radio, let it answer, and read the answer as a picture.