Unencrypted login | home

Program Information

Patient-Specific Dosimetry for CT Examinations in the Pediatric Population

D Carver

D Carver*, S Kost, J Clark, R Price, D Pickens, M Stabin, Vanderbilt Medical Center, Nashville, TN

SU-C-217A-4 Sunday 1:30:00 PM - 2:15:00 PM Room: 217A

Purpose: Radiation-dose estimates from CT scans are commonly reported as CTDI values (mGy), which are derived from standard measurements made with one of two standard (16 or 32 cm diameter) acrylic phantoms for a given technique (kVp, mAs, collimation, and pitch). Values are assigned uniformly across pediatric and adult patients without regard to specific patient size and weight. The goal of this research is to create a database of whole body and organ doses for clinical CT scans in pediatric subjects to establish more patient-specific dose estimates. Another goal of this project is to investigate how marrow dose and risk from CT imaging varies with age.

Methods: We developed a Monte Carlo radiation-transport code to simulate different CT scans. Several pediatric subjects that received chest, abdomen, and pelvis scans were selected for segmentation into voxelized phantoms for importation into our code. A series of anthropomorphic NURBS (non-uniform rational b-spline) phantoms were developed to realistically represent median, as well as larger and smaller normal weight and obese pediatric individuals of various ages. With the development of a large number of these phantoms, dose estimates for individual pediatric subjects can be assigned from the most similar phantom in the database.

Results: Our models provide individual organ doses from different CT exams, thus effective doses may be more accurate over those from CTDI calculations in two reference models, and future changes in organ-weighting factors can be taken into account. Comparison of organ and marrow doses for standard NURBS models and real patients of similar stature segmented manually show very good agreement.

Conclusion: Patient-specific dosimetry may be achieved through the use of patient-matched NURBS models for pediatric CT scans. The quality of dose data to be assigned to pediatric patients receiving CT examinations may be significantly improved using the information obtained in this study.

Contact Email