CT Dose Metrics (CTDI, DLP)
- CTDIvol tells you how intensely a slice was irradiated — think "dose per inch of patient."
- DLP multiplies that intensity by how long a stretch you scanned — "intensity times length."
- Both numbers describe the scanner's output, not the dose any particular patient actually absorbed.
- To estimate real patient risk you multiply DLP by a body-region conversion factor to get effective dose, in millisieverts (mSv).
- A phantom is not a person: a 350-pound adult and a toddler can get the identical CTDIvol and absorb wildly different doses.
Every CT scan hands you a little receipt on the dose page — CTDIvol, DLP, a few units you half-recognize — and the very natural instinct is to nod sagely and scroll past it. Let's actually read the receipt. It turns out the two big numbers are almost embarrassingly intuitive once you stop letting the acronyms intimidate you.
CTDI: how hard the beam hit each slice
CTDI stands for Computed Tomography Dose Index, which is a mouthful that means roughly "how much radiation a single slice's worth of scanning delivers." Picture a lawn sprinkler doing one rotation. CTDI is asking: how heavily did it soak the patch of grass right underneath it?
The catch is that the sprinkler doesn't water just its own patch — the spray feathers out onto the neighbors. So CTDI is measured in a plastic cylinder called a phantom (a stand-in for a body) using a long pencil-shaped ion chamber that catches both the direct beam and the scatter spilling in from adjacent rotations.
The version you'll actually see reported is CTDIvol ("vol" for volume), measured in milligray (mGy). It accounts for pitch — how tightly the scan's helical turns are packed together. Crank the table speed so the turns spread apart, and each spot gets less overlapping spray, so CTDIvol drops. The thing to hold onto: CTDIvol is an intensity. It does not care whether you scanned one centimeter or the whole torso.
There are two standard phantom sizes — a smaller one used for heads and a larger one for bodies. Same scanner settings reported against different phantoms give different CTDIvol numbers, which is exactly why "head CTDIvol" and "body CTDIvol" aren't apples to apples.
DLP: intensity times the length of the trip
If CTDIvol is "how hard," DLP — Dose-Length Product — is "how hard, for how far." You take the intensity (CTDIvol, in mGy) and multiply by the scan length (in cm), landing on units of mGy·cm.
The garden-hose version: CTDIvol is the water pressure; DLP is the pressure times how long a strip of driveway you hosed down. Scan a stubby little area at high intensity, or a long region at modest intensity, and you can end up at the same DLP from opposite directions.
| Metric | Plain-English meaning | Units | Depends on scan length? |
|---|---|---|---|
| CTDIvol | Beam intensity per slice (vs. a phantom) | mGy | No |
| DLP | Intensity × how much of the body you scanned | mGy·cm | Yes |
| Effective dose | Whole-body risk-equivalent estimate | mSv | Yes (via region factor) |
The honest part: these describe the machine, not your patient
Here's the trap, and it's the one worth tattooing on the inside of your eyelids. CTDIvol and DLP describe what the scanner put out into a plastic phantom. They are not the dose a specific human body absorbed.
Why? Because a phantom is a uniform plastic cylinder, and people are emphatically not. A petite patient and a large patient scanned with the same CTDIvol absorb very different doses — the smaller body, with less tissue to soak up and scatter the beam, generally absorbs more for the same machine output. This is doubly important in kids, which is its whole own conversation over in pediatric dose.
Do not quote CTDIvol or DLP to a worried patient as "your radiation dose." They aren't. They're the scanner's output against a standard phantom. The closest thing to a personal dose is the effective dose estimate below — and even that is a population approximation, not a personal odometer.
From DLP to effective dose (mSv)
To get something that approximates real-world risk, you convert DLP into effective dose, measured in millisieverts (mSv). The quick-and-dirty method multiplies the exam's total DLP by a published conversion coefficient (often written k) that's specific to the body region — a chest factor differs from a head factor, because the same beam over the radiosensitive chest carries more risk-weight than over the relatively hardy skull.
So the recipe, conceptually:
effective dose (mSv) ≈ DLP (mGy·cm) × k (region factor)
The whole point of effective dose is to put a head CT, a chest CT, and a background year of cosmic rays onto one comparable risk scale. I won't quote specific k values here because they depend on the reference data and patient age you're using — look them up rather than trusting a number you half-remember. The connection between dose and actual harm is the domain of radiation biology and risk.
When you tweak a protocol to lower dose — dropping tube current, raising pitch, or letting the scanner modulate output on the fly — CTDIvol is your real-time scoreboard. Watching it move is how the ALARA principle stops being a poster on the wall and becomes a knob you actually turn.
Putting it together
CTDIvol is intensity. DLP is intensity stretched over the scan length. Effective dose is DLP filtered through a region-specific factor to approximate human risk in mSv. The metrics live downstream of everything in CT physics and reconstruction, and upstream of every "is this scan worth it?" conversation.
The single most useful habit: read the dose report on your own scans, watch how CTDIvol shifts when the patient's size changes or you adjust a protocol, and never confuse the scanner's receipt with the dose a real body absorbed.