A friend asked me to tell him about computed tomography. I started to explain it — and then realized the explanation wants to be touched, not read. So I built the actual trick, running live, and it’s the most quietly astonishing thing I’ve made in a while.

Here’s the problem CT solves: a plain X-ray is a shadow. Everything the beam passes through — skin, bone, tumor, table — gets flattened into one image, stacked on top of itself. A single shadow throws away the depth. So how do you see a slice of the inside, cleanly, without cutting anyone open?

You take the shadow from hundreds of angles, and you let the overlap give you the answer. Drag the slider below — or hit ▶ Watch it resolve — and watch a formless smear tighten into a diagnosable skull.

(On a small screen, or want it full-bleed? Open the explainer on its own page — it’s a single self-contained file that runs with the wifi off.)

Everything in it is real. The middle panel is a genuine Radon transform — the 1917 math that proved you can recover a shape from all its line-integrals, decades before anyone had a machine that needed it. The reconstruction is real filtered back-projection. I checked the output against the original numerically before shipping: with the ramp filter on, the rebuilt image correlates 0.93 with the hidden truth; with it off, 0.57 — a foggy blur. That gap is the whole reason CT is useful and not just clever, and you can toggle it and see it for yourself.

The idea underneath is the one I keep circling: CT is an inverse problem — recovering a hidden interior from measurements taken only on the outside. The same math that images a brain tumor also maps the molten inside of the Earth from earthquake waves, and the gas swirling around a black hole. It’s mathematical X-ray vision. Once you have it, you can point it at almost anything you can’t cut open.

The first clinical scanner, by the way, was built at EMI in 1971 — the Beatles’ record label, flush with Sgt. Pepper money. So, faintly, the Beatles paid for the ability to see inside a living skull. I love that the machine now imaging brains was funded by the sound of the 1960s.