Click here to


Are you sure ?

Yes, do it No, cancel

A New Method for CT Dose Estimation by Determining Patient Water Equivalent Diameter From Localizer Radiographs Using Abdomen and Chest Phantoms

C Burton*, D Beaulieu , R MacDougall , Boston Children's Hospital, Boston, MA


(Wednesday, 7/17/2019) 10:00 AM - 10:30 AM

Room: Exhibit Hall | Forum 9

Purpose: AAPM Report 220 introduced patient size-specific dose estimate (SSDE) to estimate patient-absorbed dose from CTDIvol by incorporating correction factors based on patient WED. While estimating WED from CT axial scans is possible, it cannot be used to estimate SSDE before the CT acquisition. Estimating WED from CT localizer radiographs, prior to CT axial scans, would allow for WED to be included into data-driven clinical workflows such as size adaptive protocol selection, such as diagnostic reference ranges (DRRs), for example, which provide a minimum estimated patient dose. Additional benefits include reduced data overhead if axial images are not stored and errors related to axial calculation. Recently, a calibration of localizer pixel values in terms of water attenuation was proposed. In this study we apply this method on anthropomorphic phantom data from a CT scanner.

Methods: CT axial and CT localizer scans of CTDIvol phantoms of 1.0 (PMMA rod), 10, 16, and 32 cm diameters were acquired. Localized pixel values (LPV) lines within the CT axial scan range were extracted. The calibration curve was plotted as water-equivalent area per localizer-lateral dimension as a function of LPV. CT axial and CT localizer CIRS abdomen and chest phantoms were scanned. The localizer LPV lines were used with the calibration curve to calculate WED. The CT axial-based WED as a function of localizer-based was plotted.

Results: The localizer-based WED calculated had an excellent correlation (R^2=0.98) with the axial-based approach. The minimum and maximum difference between localizer-based and gold standard was 6.6 and 14%, respectively. The correlations were excellent for abdomen only (R^2=0.99) and chest (R^2=0.98) scans with a minimum/maximum difference of 6.8/8.8% and 6.6/14% difference, respectively.

Conclusion: The calibration and model-based magnification method gives an accurate estimate of WED that is used to achieve SSDE.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email