The Impact of 4DCT Dataset Representation on Radiotherapy Dose Calculation

The purpose of this study was to determine how different scanning techniques on 4DCT images impacted the perceived target volume, and how this in turn affected the dose calculation and dose distribution to the target volume. Acquired 4DCT datasets were reconstructed to form Maximum Intensity Projections (MIP), Average Intensity Projections (AIP) and free breathing datasets. Comparisons of the resultant CT number, mass density, point dose values, and dose distribution were made for the different datasets. These comparisons showed how different scanning techniques used to manipulate the 4DCT dataset affect treatment planning and dose calculation for radiotherapy for a moving tumour volume within the lungs. While the CT numbers for the AIP, MIP and the free breathing datasets were different, the resultant density values for the moving target volume showed similar results for the MIP and free breathing images. For the MIP dataset, the target volume density ranged from 1.03 g/cm3 to 1.04 g/cm3. For the Free breathing dataset, the volume density ranged from 0.90 g/cm3 to 1.02 g/cm3. The AIP dataset showed lower target volume density, ranging from 0.50 g/cm3 to 0.86 g/cm3. Dose to a point was both measured and calculated for comparison as well. The calculated point dose on the different image types was calculated using the Pinnacle3 treatment planning system, while an ionization chamber was used to measure dose to a point in the phantom. Results showed lower calculated dose to a point for all the image types, in comparison to measured point dose values. However, a more clinically relevant assessment would be to compare the dose distribution to the target volume. Comparisons of the dose distribution for the different image types (the AIP, MIP and free breathing) compared to the dose distribution averaged across the 8 phases of the 4DCT dataset was made. Comparisons were also made with increasing magnitude of motion to identify if the results were the same as the magnitude of motion increased. As motion increased, it was found that while MIP images were acceptable for dose calculation and treatment planning, the AIP images became less suitable. Results obtained here are based on simulated motions on a simplified phantom. Although simplified geometries and motions were used in this study, the results provide some guidance when applying AIP or MIP related 4DCT treatment planning procedure for small tumours moving within a heterogeneous medium.