Size-Specific Dose Estimates

A scanner index is not a patient's absorbed dose

CTDIvol describes scanner output using a standardized reference phantom. It does not account fully for a particular patient's dimensions, composition or organ locations. Two patients scanned with the same displayed index can have different absorbed-dose distributions. The size-specific dose estimate (SSDE) adjusts the index for patient size to estimate dose in the scanned region more meaningfully.

First identify the reference phantom. A displayed CTDIvol may be based on a 16 cm or 32 cm PMMA phantom depending on the examination and system. Do not infer that every pediatric body examination uses 16 cm or that every displayed value uses 32 cm. The correct conversion factor must match the reference actually used. Comparing unlabeled indices can create an apparent dose difference caused partly by different references.

Physical dimensions and effective diameter

AAPM Report 204 provides size-conversion approaches using physical patient dimensions. Measurements may include anteroposterior (AP) and lateral dimensions or their combination. If a cross-section is approximated by an ellipse, the effective diameter is:

$$d_{\mathrm{eff}}=\sqrt{AP\times LAT}$$

For a hypothetical AP dimension of 20 cm and lateral dimension of 30 cm, effective diameter is √600 = 24.49 cm. This is a geometric approximation. Measure the relevant cross-section with the appropriate method and account for localizer magnification when a projection image is used. A hand measurement from an off-center scout is not automatically the true patient dimension.

Patient size can vary along a long scan. One diameter selected at the mid-abdomen may not represent the shoulders, lungs and pelvis equally. Use the program's specified method for a regional or longitudinal estimate, documenting how it was obtained. Physical dimensions describe size but not every tissue's attenuation behavior.

Water-equivalent diameter includes attenuation

Water-equivalent diameter (Dw), addressed in AAPM Report 220, combines cross-sectional size and attenuation. An air-containing chest and a muscular abdomen can have similar physical widths but different attenuation. Dw estimates the diameter of a water-equivalent cross-section corresponding to the integrated attenuation of the measured patient section.

For a complete patient cross-sectional ROI with area A and mean CT number H, a useful expression is water-equivalent area Aw = (1 + H/1000) × A, followed by Dw = 2√(Aw/π). The method assumes appropriate image calibration and segmentation. It should not be applied to an arbitrary small ROI in one organ as though that represented the entire patient.

For an arithmetic example, A = 500 cm² and H = −200 HU give Aw = 0.8 × 500 = 400 cm². Dw = 2√(400/π) = approximately 22.57 cm. The negative mean reflects the chosen mixed-tissue cross-section; the calculation is not a published normal chest value or a recommended fixed factor for all patients.

Select the correct size factor

SSDE = size factor × CTDIvol, with the result in mGy. Obtain the factor from the appropriate published method or validated software for the region, size metric and reference phantom. A factor based on 16 cm cannot be substituted into a 32 cm-based index. A patient diameter of 32 cm does not automatically make the factor one because patient composition and PMMA are different.

In a worked exercise, suppose the problem explicitly supplies a validated factor of 1.52 and CTDIvol of 10 mGy. SSDE is 1.52 × 10 = 15.2 mGy. The factor is a supplied teaching input, not a claimed AAPM table entry for a particular diameter. If a different appropriate factor is 0.80 for another stated case, the same 10 mGy index yields 8 mGy. The comparison illustrates why the factor and reference must be recorded.

Limits of the estimate

SSDE estimates dose in the scanned region; it is not an exact dose to every organ, the peak skin dose or an effective dose in mSv. Organ location, scan boundaries, modulation and patient composition affect actual organ dose. Do not convert SSDE directly into a personal cancer probability.

Truncation, metal, inaccurate contours or an inappropriate field of view can distort a Dw estimate. A partially imaged body contour may omit attenuation information, while an included table or support can confound segmentation. Review unusual automated estimates and identify the limitation instead of assuming that a number generated by software is inherently accurate.

Use SSDE in a practical review

Suppose a small patient has an adult-like CTDIvol but a substantially larger size factor. Review whether the acquisition was properly adjusted for size and task. The answer is not automatically to multiply every displayed value by two; it is to use the validated estimate and the actual image quality in the protocol review. Keep the result alongside the scan range and relevant technical settings so that comparisons are meaningful.

Reference: AAPM Report 220.