School of Medicine and Health Sciences Poster Presentations
Comparison of DLP-based effective dose to Monte Carlo-based effective dose in low dose chest CTs
Document Type
Poster
Keywords
Lung cancer; low dose CT; effective dose
Publication Date
Spring 2017
Abstract
Introduction: Lung cancer is very difficult to detect during its early stages, as outward symptoms are not typically expressed early in the disease process. Advances in low dose CT have made it possible to screen high-risk patients and make earlier diagnoses. It is important to strike a balance between radiation exposure and image resolution, and the recommended effective dose (ED) of radiation for these scans is 1.5 mSv, much lower than the 8 mSv dose of a typical diagnostic chest CT scan. The purpose of this study was to compare the rapid formulaic dose length product (DLP)-based method of calculating ED to the Monte Carlo-based method, which is regarded as the gold standard.
Methods: This was a HIPAA compliant retrospective study. Dose data from 85 non-contrast low dose chest CT’s used for lung cancer screening were collected. Monte Carlo simulated organ based effective dose (EDMC) was calculated using Radimetrics software, a commercially available radiation dose tracking software. The DLP-based effective dose (EDDLP-B) was calculated using the formula ED = DLP * k, where k is the conversion coefficient, which are widely published. A k value of 0.015 was used for both sexes (kB), and female and male specific k-coefficients of 0.019 (kF) and 0.011 (kM) were also used respectively. ΔED was calculated as mean EDDLP – mean EDMC; and %ΔED was calculated as (mean ΔED/mean EDMC)*100. EDMC and EDDLP were compared using Wilcoxon signed rank test (WSRT) using kB, kF and kM to calculate EDDLP. Modified Bland-Altman plots were created, comparing ΔED to EDMC, and %ΔED was also plotted against patient diameter.
Results: There was statistically significant difference between EDMC and EDDLP-B (pkB (0.015) coefficient, although this was heavily influenced by gender. EDDLP-B underestimates EDMC by a mean of 31% in women (pMC and EDDLP-B in male patients (p=0.3173). EDDLP underestimated EDMC by 13% in women when using the gender specific kF; this difference was significant (pDLP underestimated EDMC by 28% in men when using the gender specific kM; this difference remained significant (p
Conclusion: DLP-based calculation of ED using the gender-neutral k-coefficient underestimates ED by 31% in women; use of female-specific k-coefficient decreases this underestimation to 13%. This should be factored into CT protocol development of low-dose chest CT’s in women. Gender-neutral k-coefficient is adequate for DLP-based ED calculation in men.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Open Access
1
Comparison of DLP-based effective dose to Monte Carlo-based effective dose in low dose chest CTs
Introduction: Lung cancer is very difficult to detect during its early stages, as outward symptoms are not typically expressed early in the disease process. Advances in low dose CT have made it possible to screen high-risk patients and make earlier diagnoses. It is important to strike a balance between radiation exposure and image resolution, and the recommended effective dose (ED) of radiation for these scans is 1.5 mSv, much lower than the 8 mSv dose of a typical diagnostic chest CT scan. The purpose of this study was to compare the rapid formulaic dose length product (DLP)-based method of calculating ED to the Monte Carlo-based method, which is regarded as the gold standard.
Methods: This was a HIPAA compliant retrospective study. Dose data from 85 non-contrast low dose chest CT’s used for lung cancer screening were collected. Monte Carlo simulated organ based effective dose (EDMC) was calculated using Radimetrics software, a commercially available radiation dose tracking software. The DLP-based effective dose (EDDLP-B) was calculated using the formula ED = DLP * k, where k is the conversion coefficient, which are widely published. A k value of 0.015 was used for both sexes (kB), and female and male specific k-coefficients of 0.019 (kF) and 0.011 (kM) were also used respectively. ΔED was calculated as mean EDDLP – mean EDMC; and %ΔED was calculated as (mean ΔED/mean EDMC)*100. EDMC and EDDLP were compared using Wilcoxon signed rank test (WSRT) using kB, kF and kM to calculate EDDLP. Modified Bland-Altman plots were created, comparing ΔED to EDMC, and %ΔED was also plotted against patient diameter.
Results: There was statistically significant difference between EDMC and EDDLP-B (pkB (0.015) coefficient, although this was heavily influenced by gender. EDDLP-B underestimates EDMC by a mean of 31% in women (pMC and EDDLP-B in male patients (p=0.3173). EDDLP underestimated EDMC by 13% in women when using the gender specific kF; this difference was significant (pDLP underestimated EDMC by 28% in men when using the gender specific kM; this difference remained significant (p
Conclusion: DLP-based calculation of ED using the gender-neutral k-coefficient underestimates ED by 31% in women; use of female-specific k-coefficient decreases this underestimation to 13%. This should be factored into CT protocol development of low-dose chest CT’s in women. Gender-neutral k-coefficient is adequate for DLP-based ED calculation in men.
Comments
Poster to be presented at GW Annual Research Days 2017.