Encrypted login | home

Program Information

The Variation of Hounsfield Unit and Relative Electron Density Determination as a Function of KVp and Its Effect On Dose Calculation Accuracy

no image available
A Ohl

A Ohl1*, S De Boer2 , (1) University of Buffalo, Buffalo, NY (2) Roswell Park Cancer Institute, Buffalo, NY


SU-E-T-599 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

Purpose: To investigate the differences in relative electron density for different energy (kVp) settings and the effect that these differences have on dose calculations.

Methods: A Nuclear Associates 76-430 Mini CT QC Phantom with materials of known relative electron densities was imaged by one multi-slice (16) and one single-slice computed tomography (CT) scanner. The Hounsfield unit (HU) was recorded for each material with energies ranging from 80 to 140 kVp and a representative relative electron density (RED) curve was created. A 5 cm thick inhomogeneity was created in the treatment planning system (TPS) image at a depth of 5 cm. The inhomogeneity was assigned HU for various materials for each kVp calibration curve. The dose was then calculated with the analytical anisotropic algorithm (AAA) at points within and below the inhomogeneity and compared using the 80 kVp beam as a baseline.

Results: The differences in RED values as a function of kVp showed the largest variations of 580 and 547 HU for the Aluminum and Bone materials; the smallest differences of 0.6 and 3.0 HU were observed for the air and lung inhomogeneities. The corresponding dose calculations for the different RED values assigned to the 5 cm thick slab revealed the largest differences inside the aluminum and bone inhomogeneities of 2.2 to 6.4% and 4.3 to 7.0% respectively. The dose differences beyond these two inhomogeneities were between 0.4 to 1.6% for aluminum and 1.9 to 2.2 % for bone. For materials with lower HU the calculated dose differences were less than 1.0%.

Conclusion: For high CT number materials the dose differences in the phantom calculation as high as 7.0% are significant. This result may indicate that implementing energy specific RED curves can increase dose calculation accuracy.

Contact Email: