ABSTRACT
The results of a recent study on the limiting uncertainties in the measurement of photon radiation dose with MOSFET dosimeters are reported. The statistical uncertainty in dose measurement from a single device has been measured before and after irradiation. The resulting increase in 1/f noise with radiation dose has been investigated via various analytical models. The limit of uncertainty in the ubiquitous linear trend of threshold voltage with dose has been measured and compared to two nonlinear models. Inter-device uncertainty has been investigated in a group of 40 devices, and preliminary evidence for kurtosis and skewness in the distributions for devices without external bias has been observed.
Subject(s)
Particle Accelerators , Radiometry/methods , Radiotherapy Planning, Computer-Assisted/methods , Radiotherapy/methods , Algorithms , Calibration , Dose-Response Relationship, Radiation , Humans , Lipids , Models, Statistical , Models, Theoretical , Monte Carlo Method , Normal Distribution , Photons , Radiation , Radiation Dosage , Radiotherapy Dosage , Reproducibility of Results , Sensitivity and Specificity , SoftwareABSTRACT
The feasibility of large-geometry Metal Oxide Semiconductor Field Effect Transistor (MOSFET) devices has been assessed for both active and passive neutron dosimetry and use in radiotherapy environments. Neutron sensitivity has been enhanced with the use of polymeric cement surrounding the gate region. Neutron activation via nuclear interaction processes is a potential problem with conventionally packaged and fabricated devices. To overcome this problem, a unique low-activation device design is described. Standard Dual in-Line devices, modified with polymeric cement and boron loaded cement have been exposed to gamma rays (60Co) and neutrons (gamma-ray shielded 252Cf) to provide neutron sensitivity estimates. The results show that the neutron sensitivity can be increased by a factor of approximately three by the use of a thin layer of polymeric cement over the gate region. Essentially zero activation is observed in the activation-reduced design compared with 1000 cps in the conventional design MOSFET when both are exposed under identical conditions to a neutron field from a gamma-ray shielded 252Cf isotopic source.
