Gamma Camera & SPECT
- A gamma camera is a flat detector that maps where gamma rays are coming out of the patient — the radiotracer is the light source, and the body is the thing glowing.
- The collimator is a slab of lead with holes that throws away every gamma ray traveling the "wrong" way, so only straight-line photons get counted. This is what gives you a picture instead of a blur.
- A scintillation crystal (usually sodium iodide) turns each gamma ray into a tiny flash of light, and photomultiplier tubes behind it figure out where that flash happened.
- Planar imaging gives you a flat shadow; SPECT spins the camera around the patient to reconstruct a 3D volume, just like CT does with X-rays.
- Resolution and sensitivity are always fighting each other — make the collimator pickier and the image gets sharper but dimmer.
Most imaging shines energy into a patient and measures what comes out the far side. Nuclear medicine flips the whole thing on its head. Here, the patient is the lightbulb. We've injected a radiotracer that quietly emits gamma rays from inside, and the gamma camera's only job is to stand outside and photograph the glow. (For where that glow comes from, see radioactive decay and radiopharmaceuticals.) It's less "X-ray vision" and more "watching a firefly through a window in the dark."
The collimator: a lead bouncer with a strict door policy
Here's the problem. Gamma rays fly out of the patient in every direction, like sound leaving a stadium. If you just held up a detector, every point would be hit by photons from everywhere, and you'd get a uniform gray smear — no picture at all.
The fix is the collimator: a thick plate of lead drilled with thousands of long, narrow, parallel channels. Think of looking at the world through a tightly-packed bundle of drinking straws. A gamma ray traveling straight down a straw makes it through to the detector. A gamma ray coming in at an angle slams into a lead wall and gets absorbed. So only photons that were traveling perpendicular to the camera survive, which means each spot on the detector can be traced back to the spot in the patient directly in front of it.
This is the same logic as the grids and collimation used in plain radiography — reject the off-angle photons, keep the honest ones. Different machine, same instinct: geometry is what turns scattered light into an image.
The collimator is also where the cruel trade-off lives. Make the holes longer and narrower and you reject more stray rays, so the image gets sharper — but you also throw away most of your photons, so the image gets dimmer and noisier. Loosen up to catch more counts and everything blurs. There is no free lunch; there is only the collimator you chose.
The crystal and the tubes: catching the flash
Behind the collimator sits the scintillation crystal, classically a single large slab of sodium iodide doped with a trace of thallium. "Scintillation" is a fancy word for it sparkles when hit — each incoming gamma ray deposits its energy and the crystal answers with a faint flash of visible light. One gamma in, one tiny twinkle out.
Those twinkles are far too dim to see, so behind the crystal sits a grid of photomultiplier tubes (PMTs) — light amplifiers that take a few photons and spit out a measurable electrical pulse. Crucially, the tubes nearest the flash see it brightest. By comparing how strongly each tube responds, the camera's electronics triangulate the exact spot the flash happened. That position logic is the heart of the gamma camera.
Energy windowing: ignoring the photons that lie
Not every gamma ray that reaches the crystal is trustworthy. Many have scattered inside the patient — they ricocheted off tissue, changed direction, and lost a little energy along the way. If we counted those, they'd place a flash in the wrong location and fog the image.
The escape hatch is that scattered photons come in with less energy than the originals. The camera sets an energy window around the tracer's known gamma energy and only accepts photons whose energy lands inside it. A photon that shows up underpowered gets rejected as a probable scatterer. It's a bouncer checking IDs: right energy, you're in; suspiciously low, you're out.
Set the energy window wrong and the image quietly degrades — too wide and you let scattered photons back in (hazy, low-contrast image); too narrow and you reject good counts (noisy, grainy image). The picture can look "fine" while being subtly off, which is why technique and quality control matter as much as the hardware.
From flat shadow to 3D: SPECT
A single gamma-camera view is planar imaging — a flat projection, like a shadow on a wall. It's quick and useful, but everything in front of and behind your target is squashed into one plane, so a hot spot deep in the body can hide behind one nearer the surface.
SPECT — single-photon emission computed tomography — fixes this the same way CT fixes the plain radiograph. The camera (often two or three heads) rotates around the patient, grabbing planar projections from many angles. A reconstruction algorithm then back-calculates the 3D distribution of tracer that must have produced all those views. Suddenly you can scroll through slices and put a finding exactly where it lives instead of guessing its depth.
Bolting a CT scanner onto the same gantry gives you SPECT/CT — the functional "where is the tracer glowing" map laid precisely over the anatomical "what structure is that" map. The CT also helps correct for attenuation, since gamma rays from deep inside the body get partly absorbed before they ever escape.
The one thing to carry out the door
A gamma camera is fundamentally a position-finding machine for light flashes: a collimator decides which gamma rays count, a crystal turns them into sparkles, and tubes pin down where each sparkle was. SPECT just spins that camera in a circle to recover depth. Get comfortable with this and the photon-counting cousin, PET, will feel like a natural next step rather than a whole new language.