Premium
SU‐FF‐T‐152: Convolution/superposition Algorithm and High‐Z Dental Materials: Dosimetric Study in a Solid Water Slab Phantom
Author(s) -
Spirydovich S,
Papiez L,
Langer M,
Thai V
Publication year - 2006
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.2241076
Subject(s) - superposition principle , convolution (computer science) , imaging phantom , dosimetry , algorithm , physics , monte carlo method , materials science , mathematics , optics , nuclear medicine , mathematical analysis , computer science , statistics , medicine , artificial intelligence , artificial neural network
Purpose To address the accuracy of the dose calculated with convolution/superposition algorithm in the presence of high‐Z dental materials. Methods and Materials Three methods were utilized to access the dose: convolution/superposition algorithm, Fluence Map Monte Carlo (FMMC) method, and radiochromic film. We considered a solid water® slab phantom which had an embedded high‐Z material. For dose calculations and measurements we used a 6MV photon beam from a clinically commissioned linear accelerator. Results We observed a close agreement for the dose measured with radiochromic film and the dose calculated with FMMC algorithm. On the other hand, a large discrepancy was discovered for the dose calculated with the convolution/superposition algorithm compared to the dose obtained with measurement or FMMC algorithm. The greatest discrepancy was observed downstream from the high‐Z cerrobend inhomogeneity where the convolution/superposition algorithm calculated a dose which was higher than the dose measured with radiochromic film by 10–20% depending on the size of and the distance from the inhomogeneity. Clinically this finding shows that the delivered dose would be 10– 20% less than the prescribed dose which was calculated with convolution/superposition algorithm. In the region upstream from all the studied high‐Z inhomogeneities the convolution/superposition algorithm was underestimating the delivered dose. The convolution/superposition algorithm was unable to properly estimate the dose enhancement due to the increased backscatter near the inhomogeneity. Conclusions The convolution/superposition algorithm may significantly overestimate the actual dose in the site of the tumor located downstream from the high‐Z dental restorations or prostheses.