Loading…

SU‐E‐T‐70: Are the Dose Response Functions of Ionization Chambers Gaussian Or Non‐Gaussian?

Purpose: Experimental results from independent investigations have shown that the dose response functions K(x) of a large variety of ionization chambers can be characterized by Gaussian functions. Theoretical considerations and Monte‐Carlo simulations, however, have revealed that the K(x) are non‐Ga...

Full description

Saved in:
Bibliographic Details
Published in:Medical Physics 2013-06, Vol.40 (6), p.219-219
Main Authors: Looe, HK, Harder, D, Poppe, B
Format: Article
Language:English
Subjects:
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Purpose: Experimental results from independent investigations have shown that the dose response functions K(x) of a large variety of ionization chambers can be characterized by Gaussian functions. Theoretical considerations and Monte‐Carlo simulations, however, have revealed that the K(x) are non‐Gaussian. In this work, experimental and Monte‐Carlo studies were compared to understand the origin of the general occurrence of Gaussian kernels for ionization chambers with different geometries, using Fourier analysis. Methods: The measured dose profile M(x) is related to the true dose profile D(x) via the convolution M(x) = D(x)*, K(x). The best G‐values of K(x) for three ionization chambers with different geometries: (i) cylindrical; (ii) square; and (iii) parallel‐plate were determined experimentally by comparing M(x) with D(x)*, K(x). Monte Carlo simulations of the fluence response kernels KM(x) were performed by stepwise shifting a slit beam across each chamber. The function K(x) of each chamber is obtained by deconvolving KM(x) with the dose deposition kernel KD(x). Results: Our experiments show that the K(x) of the investigated ionization chambers can be all characterized by Gaussian functions, regardless of their geometries. However, the true K(x) obtained with Monte‐Carlo simulations are non‐Gaussian, exhibiting increased response at the positions of the inner electrode and the chamber wall. Fourier analysis of the true K(x) and its Gaussian approximation G(x) has revealed that FT[G(x)] closely resembles FT[K(x)] at spatial frequencies below about 0.15 mm−1. Furthermore, Fourier transforms of a 2 cm wide dose profile convolved with G(x) or K(x) are not showing noticeable differences. Conclusion: The Gaussian kernel G(x) experimentally observed are approximations adequately characterizing the spatial resolution of ionization since clinical dose profiles have only negligible components at frequencies higher than 0.1 mm −1 due to physical and geometrical reasons.
ISSN:0094-2405
2473-4209
DOI:10.1118/1.4814505