Unencrypted login | home

Program Information

Position Sensitivity of Calculated Dose-Length-Product-To-Effective-Dose Conversion Factors in Computed Tomography Examination

P Caracappa

Y Gao, A Ding, P Caracappa*, X Xu, Rensselaer Polytechnic Institute, Troy, NY

SU-C-217A-3 Sunday 1:30:00 PM - 2:15:00 PM Room: 217A

Purpose: To quantify the sensitivity of Dose Length Product to Effective Dose conversion factors (k-factors) to scan length and position.

Methods: To convert the recorded Dose Length Product (DLP) to an estimate of Effective Dose (E), a set of conversion factors, known as k-factors, are used. These published k-factors were developed for a limited set of standard protocols with fixed scan lengths, and most are based on stylized models of the human body. Using realistic models of male and female (RPI-AM and RPI-AF) and a validated model of the CT source, Monte Carlo simulations are preformed to calculate the dose to each organ on a slice-by-slice basis. For each slice, the organ doses are converted into 'local' k-factors. Over a full scan length, the effective doses from the individual slices and the DLPs are summed, so the k-factor for the scan is equivalent to the average of the local k-factors in the scan range.

Results: Simulations of 1-cm thick CT slices were performed over the entire range of the torso for the RPI-AM and RPI-AF phantoms. The local k-factors calculated range from a minimum of 0.0128 in the upper shoulder region (above most of the radio-sensitive organs of the chest), to a maximum of 0.0369 in the region around the female gonads. The calculated k-factor for a standard chest scan was 0.0289, which compares favorably to published values. However, the k-factor decreases as the scan range is expanded. Consequently, a chest scan that is only 10 cm longer would have a k-factor 18% lower when compared to the standard protocol.

Conclusion: By improving the understanding of the sensitivity of k-factors to scan length and position, more accurate estimates of effective dose from CT examination can be provided without sacrificing the efficiency and simplicity that k-factors provide.

Contact Email