Author: Renxin Chu, Jelena Mihailovic, Kai Yang, Lifeng Yu, Da Zhang 👨🔬
Affiliation: Massachusetts General Hospital, Mayo Clinic, Boston Children's Hospital, UVA Health University Hospital, University of Missouri Health Care 🌍
Purpose:
A frequently used method to estimate effective dose (ED) in CT is by multiplying the dose length product (DLP) by a conversion factor, known as k-factor. The k-factor in this method is susceptible to variations associated with many scans and patient related factors. The purpose of this study was to estimate the uncertainties associated with different scanning parameters and patient characteristics.
Methods:
The k-factors were obtained from ED and DLP estimates using Monte Carlo simulation-based software (VirtualDose CT, Virtual Phantoms Inc.) for 25 computational phantoms representing different patients [gender, age, size (represented by body-mass-index, BMI)], three different CT scanner manufacturers (Siemens, GE, Phillips) and seven commonly performed CT exams. For each exam type, volume CT dose index (CTDIvol) was fixed at 10 mGy with varying kV (80-120) and scan length (±5 cm) from regular settings. Multivariable linear regression method was used for empirical formula for k-factors=a×kV+b×BMI+c×Gender+d×ScanLenght+e×Vendor and for analysis of the impact of changes as uncertainty (%) = coefficient×variable range/mean k-factor.
Results:
k-factors for all exams increased with tube voltage, with the highest dependence for chest and abdomen for adult, chest, abdomen, AP and CAP for pediatric. Negative dependence of k-factors on BMI was observed with ~13% lower values for male vs female. However, that difference was not seen in pediatric phantoms. The same trend followed in all three CT manufacturers. Analysis showed that the uncertainties of k-factors were influenced most by scan length and BMI for adults, and age and scan length for pediatric phantoms. Even though the same parameter changes were investigated in all manufacturers, there was a difference in uncertainties between them for the same CT exam.
Conclusion:
Our study showed that large k-factors uncertainties were influenced by BMI and scan length and age affecting ED estimation through the k-factor method in clinical settings.