Effective Dose for Computed Tomography in Large, Clinical Populations

Conference Paper

PURPOSE Effective dose can efficiently integrate multiple organ dose values into a singular scalar value of radiation dose. The effective dose can also be calculated using dose length product (DLP) to effective dose conversion coefficients. The purpose of this study was to compare the results from these two methods across a large clinical population. METHOD AND MATERIALS This IRB-approved study included 579 chest and 247 abdominopelvic exams from two scanners. The organ doses were estimated based on an established Monte Carlo method. Each patient was matched to an XCAT anthropomorphic phantom of the same anatomical height. Precomputed Monte Carlo data were then applied along with adjustments for tube current modulation and patient size. A bias correction factor was also applied to each organ dose to calibrate the organ doses to validated ground truth dose values. The effective dose of each exam was calculated using both dose-length-product based (EDk) and organ-dose-based (EDOD) methods. The EDk was estimated by extracting the DLP and DLP-to-effective dose conversion coefficients defined by ICRP 102. The EDOD was derived using the patient-informed organ doses from the exam using ICRP 103 weighting factors. The EDk was compared to EDOD and CTDIvol. RESULTS For the abdominopelvic protocol, EDk was > EDOD by 80.2±48.7% on average (min: -28%; max: 235%; median: 75%). For the chest protocol, EDk was > EDOD by 64.1±33.1% on average (min: -18%; max: 191%; median: 60%). EDk was highly influenced by mean CTDIvol (abdominopelvic: R=0.9692, chest: R=0.9557). EDk was higher than EDOD, especially for larger patients (abdominopelvic: 9.6%, 66.2%, and 115.6% higher; chest: 30.2%, 77.9%, and 116.2% higher on average for patients with 22, 30, and 35 cm water equivalent diameters, respectively). This can be attributed to the contribution of doses absorbed by remainder tissue in EDk calculations. CONCLUSION In this work, we compared the effective dose calculated using DLP-based methods versus those based on organ dose. The result showed that DLP methods may over-estimate effective dose, especially for larger patients CLINICAL RELEVANCE/APPLICATION This clinical cohort study showed the substantial differences of effective dose estimated using DLP vs. organ-dose-based methods. This effect is more prominent for larger patients. The study indicates that radiation dose monitoring and reporting could benefit from accurate and robust organ dose estimation.

Duke Authors

Cited Authors

  • Fu, W; Ria, F; Wilson, J; Kapadia, AJ; Segars, WP; Samei, E

Published Date

  • November 27, 2018

Conference Name

  • RSNA 104th Scientific Assembly and Annual Meeting

Conference Location

  • Chicago (IL)

Conference Start Date

  • November 25, 2018

Conference End Date

  • November 30, 2018