| Dosimeter performance under varying photon energy and dose fractionation were investigated using variance-aware statistical modeling with regulatory accreditation criteria.
A multifactorial experimental design was implemented to evaluate three dosimeter technologies (Direct Ion Storage (Instadose VUE), Optically Stimulated Luminescence (OSL), and
Thermoluminescent Dosimeter (TLD)) across three photon energies (20 keV, 206 keV, and 662
keV) and six fractionated irradiation schedules over a 30-day period with equal cumulative dose
(42 mSv).
Performance was assessed using generalized least squares (GLS) modeling to account for
heteroscedasticity and interpreted within the ANSI/HPS N13.11-2022 bias–precision framework.
Results demonstrated that dosimeter performance is intrinsically technology-dependent and
strongly influenced by photon energy, with pronounced response divergence at low energy (20
keV). Instadose VUE performed comparably to OSL and TLD at 206 keV and 662 keV but exhibited greater deviation at 20 keV. Instadose VUE and OSL exhibited stability under fractionated
irradiation; however, a localized three-way interaction was observed for TLDs at 20 keV, consistent with trap kinetics intrinsic to thermoluminescent materials.
ANSI/HPS N13.11-2022 evaluation shows that TLD failed deep-dose performance due to
increased variability at low photon energies, whereas OSL and Instadose VUE met deep-dose criteria. Shallow-dose Hp (0.07) results show compliance for OSL and TLD, while Instadose VUE
did not, reflecting design optimization for deep-dose measurement and limitations in low-energy
response.
This thesis establishes that dosimeter reliability is governed by the structural stability of
detector response functions under varying photon energies, rather than being determined solely by
calibration accuracy.
Keywords: Direct Ion Storage Dosimeter, Multifactorial Experimental Design, Photon Energy
Dependence, Dose Fractionation, Generalized Least Squares, ANSI/HPS N13.11-2022. |