Grids, Scatter & Collimation
- Scatter is X-ray photons that get knocked off course inside the patient and arrive at the detector pointing the wrong way — it's fog that washes out contrast.
- Collimation shrinks the beam to just the body part you care about, which makes less scatter in the first place (and lowers dose). Always your first move.
- A grid sits between the patient and the detector and absorbs scatter that arrives at the wrong angle, but it costs you dose because you have to crank up the technique to compensate.
- Big body part + big beam = lots of scatter; tiny part (a finger) = barely any, so you skip the grid entirely.
Imagine taking a photo through a foggy windshield. The thing you want to see is still out there, but everything's washed out and gray because stray light is bouncing around between you and it. That fog is the enemy of every radiograph, and it has a name: scatter. This page is about where the fog comes from and the two tools we use to fight it — squeezing the beam down (collimation) and installing a tiny set of venetian blinds in front of the detector (a grid).
Where scatter comes from
When an X-ray photon dives into a patient, a few things can happen. Some photons sail straight through and hit the detector exactly where they're supposed to — those are the good ones carrying your image. Some get absorbed entirely. And some get deflected: they ricochet off an atom, change direction, and keep going at a new angle (the physicists call this Compton scatter, and at diagnostic X-ray energies it's the main source of scatter). This builds on how the beam interacts with tissue in the first place — if that feels shaky, detour through attenuation and radiographic contrast.
The problem is that a scattered photon still lands on the detector — just in the wrong spot. It doesn't carry information about where it came from anymore; it's a liar. Pile up millions of these liars and you get a uniform gray haze layered over your image, dragging down the difference between light and dark. That difference is your contrast, so scatter is contrast's mortal enemy.
More body to plow through means more scatter. That's why a lateral lumbar spine or an abdomen is a scatter factory, while a hand or a baby's chest is practically scatter-free. Thickness and field size are the two biggest dials.
Collimation: shrink the beam first
The cheapest, smartest move is to not make the fog in the first place. Collimation means closing down the adjustable lead shutters in the tube so the beam only illuminates the anatomy you actually need. Picture narrowing a garden hose nozzle from "wide spray" down to a focused stream.
Two things happen at once when you collimate tightly. First, less tissue gets irradiated, so less scatter is generated — better contrast for free. Second, you've shrunk the dose to everything outside the field, which is a gift to the patient. This is the same ALARA logic that runs through all of ALARA and protection principles: don't irradiate what you don't need to see.
Tight collimation is the rare free lunch in radiology — it improves image quality and lowers dose at the same time. If you only remember one scatter-reduction trick, remember to collimate.
The grid: venetian blinds for photons
When collimation isn't enough — say, a thick adult abdomen — we add a grid: a thin panel of alternating strips of lead and a low-density spacer, sitting between the patient and the detector. The lead strips are angled so that photons traveling straight from the tube slip neatly between them, while photons arriving at a cockeyed angle (the scattered liars) slam into a lead strip and get absorbed.
It's exactly like venetian blinds tilted toward the sun: light coming from the right direction passes through; light coming sideways gets blocked.
The catch is honesty: the grid can't tell a useful photon from a scattered one by intent, only by angle. It throws away some good photons too, plus it blocks the scatter you were trying to lose. So to get a properly exposed image, you have to increase the technique (more mAs) — which means more dose to the patient. A grid is a contrast-for-dose trade. We make that trade when the scatter problem is bad enough to be worth it.
| Situation | Use a grid? | Why |
|---|---|---|
| Adult abdomen, lumbar spine, chest | Yes | Thick part, lots of scatter, contrast would be ruined without one. |
| Extremities (hand, foot, wrist) | No | Thin part makes little scatter; the dose penalty isn't worth it. |
| Small pediatric / neonatal exams | Often no | Tiny body, minimal scatter, and we're extra protective of dose. |
The gotchas that bite everyone
Grids are picky about geometry, and lining them up wrong creates its own artifacts.
Most grids have their lead strips angled to match a specific focusing distance and a specific center. Common grid errors — putting the grid in upside down, off-center, tilted, or shooting from the wrong distance — cause grid cutoff, where the lead strips start absorbing the good primary photons too. Because fewer photons reach the detector, the image comes out underexposed — too light across the whole field, or lighter on one side. The fix is mechanical, not a technique change: line the grid up correctly.
Collimate first, grid second. Collimation prevents scatter and lowers dose; a grid only cleans up the scatter that's left, and it costs dose to do so.
So the whole strategy is just fog management. Squeeze the beam to make less fog, then, if the body part is thick enough to fog up anyway, slot in the venetian blinds to catch what's left — and accept that you'll pay for the cleaner picture in a little extra dose. The art is knowing when that picture is worth the bill, and it shows up everywhere in resolution, noise and contrast.