CTDIvol and DLP

Two numbers appear on every CT console before you scan. CTDIvol describes the intensity of the protocol in a standard phantom; DLP multiplies it by scan length. Neither is the patient's dose, and knowing why is the start of optimisation.

By the ScanFlo Editorial Team · Updated 3 October 2026

CTDIvol (volume CT dose index, in mGy) is the average radiation dose a CT protocol delivers within the scanned volume of a standard acrylic phantom, 16 cm for the head and 32 cm for the body. DLP (dose-length product, in mGy·cm) is CTDIvol multiplied by the scan length, so it describes the whole acquisition. Neither is the patient's own dose: they are standard indices for comparing protocols with each other and with diagnostic reference levels.

How CTDIvol is measured

A physicist puts a 100 mm pencil ionisation chamber into holes in an acrylic (PMMA) cylinder, 16 cm across for head protocols and 32 cm for body protocols, and makes one axial rotation. The reading in the centre and at the edge gives CTDI₁₀₀. Dose is higher at the edge than in the middle of a body phantom, so the two are combined as a weighted average, CTDIw, with one third for the centre and two thirds for the periphery (AAPM Report 96).

Real scans are helical, so pitch matters. Dividing by pitch gives CTDIvol: overlap (pitch below 1) raises it, gaps (pitch above 1) lower it. AAPM Report 96 describes CTDIvol as the average dose over x, y and z within the scan volume for a phantom of similar attenuation, and stresses that it "does not represent the average dose for objects of substantially different size, shape, or attenuation". Modern scanners display it before you scan (IAEA handbook, chapter 22).

CTDIvol vs DLP

AAPM Report 96 gives the classic illustration: an abdomen-only CT and an abdomen-and-pelvis CT can have the same CTDIvol, but the second has a larger DLP because it covers more of the patient.

Worked example: a routine adult chest

Suppose the body phantom gives CTDI₁₀₀ of 6 mGy in the centre and 12 mGy at the periphery for your chest technique. CTDIw = ⅓ × 6 + ⅔ × 12 = 2 + 8 = 10 mGy. Scanned at a pitch of 1.25, CTDIvol = 10 ÷ 1.25 = 8 mGy. Cover 35 cm from lung apices to bases and DLP = 8 × 35 = 280 mGy·cm.

Compare that with the UK national DRL for a chest CT for lung cancer: 8.5 mGy and 290 mGy·cm (UKHSA NDRLs). The protocol sits just under both. For a rough population-level effective dose, multiply DLP by the adult chest coefficient from AAPM Report 96 Table 3: 280 × 0.014 ≈ 3.9 mSv. These are illustrative numbers chosen to show the arithmetic, not measurements from a particular scanner.

Diagnostic reference levels

A diagnostic reference level is a benchmark, not a limit. ICRP Publication 135 sets out how DRLs are built from survey data, usually the 75th percentile of the median doses from many facilities, and how a department should use them: if your typical value for a standard patient is above the DRL, investigate. National values differ, so use the set for your country where one exists.

In the United States, Kanal and colleagues used 1,310,727 examinations from 583 facilities in the ACR Dose Index Registry to set DRLs and achievable doses for the 10 commonest adult CT examinations, by patient size (Radiology, 2017). In India, published CT figures come mainly from regional surveys. A seven-hospital study in Kerala found a 75th-percentile CTDIvol of 68.1 mGy for single-phase head CT (Varghese et al., 2018), and a 2024 review found that Indian DRLs vary considerably between states (Malik et al., 2024).

Patient size and effective dose

Because CTDIvol describes a phantom, it misleads most for patients who are much smaller or larger than the phantom. AAPM Report 204 introduced the size-specific dose estimate (SSDE), which scales CTDIvol by a factor based on the patient's measured dimensions. It warns that, for small children, reading a 32 cm-phantom CTDIvol as patient dose can underestimate it by a factor of 2 to 3.

From AAPM Report 96, Table 3. Adult head and neck and all paediatric coefficients assume the 16 cm phantom; adult body values assume the 32 cm phantom. Effective dose estimated this way is useful for comparing examinations and explaining relative risk. It is not a measure of any one patient's risk.

What moves the numbers

Dose and mAs, pitch and effective mAs are from AAPM Report 96; the square-root noise law from the IAEA handbook, §4.5.1. Tube voltage is harder to summarise in one line, because a lower kV lowers output but changes contrast and noise in a way that depends on patient size; see CT protocol basics.

Common mistakes

• Calling CTDIvol the patient's dose. It is a phantom index; use SSDE for size, and effective dose estimates only for comparisons.
• Comparing a head CTDIvol (16 cm phantom) with a body CTDIvol (32 cm phantom).
• Judging a scan by CTDIvol alone. A protocol with a modest CTDIvol and an over-long range can still have a high DLP.
• Forgetting that a multiphase study's DLP is the sum of its phases.
• Treating the DRL as a limit or a target. It flags protocols to review; a well-optimised protocol is usually below it.
• Reading a child's dose from a body protocol's CTDIvol without size correction.