How Images Are Made (Primer)
- Every medical image is a map of how something — energy or signal — interacted with the body's tissues.
- X-rays and CT use a beam that gets partly "eaten" on the way through; dense stuff (bone, metal) eats the most and shows up white.
- Ultrasound is an echo: it sends sound in and listens for what bounces back.
- MRI doesn't use radiation at all — it nudges the body's own water and listens to it hum back.
- Different physics means different superpowers: each modality is good at seeing different things, which is the whole reason we have more than one.
Before you learn to read any study, it helps to know one thing: a medical image is not a photograph. Nobody is shining a flashlight at your liver and snapping a picture. Instead, every image is a clever way of measuring how some kind of energy interacted with your insides, then painting that interaction in shades of gray. Once you get that, the rest of radiology stops feeling like magic and starts feeling like physics with a sense of humor.
Let me walk you through the four big ways we make pictures, fastest tour in town.
X-ray: shadows on a wall
Picture making hand shadows on a wall with a flashlight. The light that gets blocked leaves a dark shape; the light that gets through lights up the wall. An X-ray works the same way, except the "flashlight" is a beam of X-rays and the "wall" is a detector behind you.
As the beam passes through, dense tissues block more of it and thin tissues let more through. The radiologists call this attenuation — in English, how much of the beam gets eaten on the way through. This is the engine behind attenuation and radiographic contrast, and it sorts everything into a handful of basic radiographic densities.
A reliable mnemonic for what shows up white: the more an X-ray beam gets stopped, the brighter the spot. Bone and metal stop a lot, so they're bright white. Air stops almost nothing, so lungs look nearly black. Everything else — fat, muscle, fluid — lives in the grays between.
The catch: an X-ray squashes a whole 3D body onto one flat image, like stepping on a layer cake and trying to read the layers. The heart, the spine, and the ribs all stack on top of each other.
CT: the layer cake, sliced
That squashing problem is exactly what CT fixes. CT is still X-rays — same beam, same attenuation — but the tube spins all the way around you, taking shadow pictures from every angle. A computer then does the heavy math to reconstruct those angles into thin cross-sectional slices, so instead of one stepped-on cake you get to look at each layer on its own.
The payoff is huge contrast detail and true 3D anatomy. The cost is that you're using more X-rays to get it, which is why dose comes up constantly in CT and almost never with a single chest film.
Ultrasound: yelling into a canyon
Ultrasound ditches radiation entirely and borrows a trick from bats and submarines: echolocation. The probe sends a pulse of sound into you, then shuts up and listens. Wherever the sound hits a boundary between two tissues, some of it bounces back. The machine times how long each echo takes to return — longer trip means deeper structure — and assembles those echoes into an image in real time.
Yell into a canyon and the time it takes to hear yourself tells you how far the far wall is; ultrasound just does that thousands of times a second.
Sound hates air and bone. It bounces almost completely off both, so anything hiding behind a gas-filled bowel loop or a rib can be frustratingly invisible. This is why ultrasound is a star for the gallbladder and a fetus (surrounded by fluid) but a poor choice for peering through the lungs.
MRI: making water hum
MRI is the strangest and most elegant of the bunch, and it uses no ionizing radiation at all. Your body is mostly water, and every water molecule has hydrogen protons that behave like tiny spinning magnets, normally pointing every which way.
Slide into the scanner's powerful magnet and those little magnets line up. A radio pulse then knocks them sideways, and as they settle back into line they release a faint radio signal — they "hum." Different tissues hum back at different rates, and the scanner translates that into a picture.
The genius of MRI is that the same patient can look completely different depending on how you time the listening. Those timing recipes are what produce T1 and T2 weighting — the knobs that make fat, water, and tissue trade places between white and dark.
That tunability gives MRI gorgeous soft-tissue detail (brains, ligaments, spinal cords), at the cost of long scan times and a magnet so strong it treats loose metal like a missile.
So why keep four of these around?
Because each one answers a different question. A summary you can lean on:
| Modality | How the image forms | Best at | Big limitation |
|---|---|---|---|
| X-ray | Beam attenuation, flattened to 2D | Bones, lungs, fast and cheap | Everything overlaps |
| CT | Spinning X-ray beam, reconstructed into slices | 3D anatomy, trauma, acute findings | Uses more radiation |
| Ultrasound | Echoes of sound pulses | Real-time, fluid, no radiation | Blocked by air and bone |
| MRI | Radio signal from nudged water protons | Soft-tissue detail, no radiation | Slow, expensive, magnet hazards |
A common beginner trap is reading "white" the same way on every study. White on an X-ray or CT means dense / blocks the beam (bone, metal, contrast). White on MRI means strong signal, which depends entirely on the sequence — fluid can be brilliant white on one image and pitch black on the next. Always know which modality you're looking at before you trust a color.
That's the whole foundation. Whenever a study confuses you, back up and ask the one question that explains everything else: what physical thing is this gray actually measuring? Answer that, and you're already reading like a radiologist.