Uncertainty calculation methods in dose assessment for dicentric chromosome assay

Area of Science:

• Radiation dosimetry
• Statistical methods in radiation biology
• Biostatistics

Background:

• Accurate estimation of Poisson confidence intervals is crucial for radiation dose assessment.
• Understanding uncertainty sources in Bayesian methods is essential for reliable results.
• Evaluating different methods for dose uncertainty coverage is vital for risk assessment.

Purpose of the Study:

• To evaluate the coverage accuracy of various Poisson confidence interval methods.
• To assess the impact of uncertainty factors on dose coverage.
• To detail uncertainty sources within the Bayesian approach.

Main Methods:

• Simulations using R-based scripts were employed to assess interval coverage.
• Exact coverage of uncertainty Poisson confidence intervals was calculated.
• Dose uncertainty interval coverage was evaluated under different conditions.

Main Results:

• The Modified Crow and Gardner method provided Poisson confidence intervals close to nominal coverage, with desirable properties.
• The unlimited simultaneous calibration interval maintained 95% coverage under parametric and nonparametric conditions, though conservatively.
• ISO propagation of errors and Bayesian methods closely approximated 95% coverage for Poisson-distributed samples.

Conclusions:

• The Modified Crow and Gardner method is recommended over the Garwood method for Poisson exact confidence intervals.
• The unlimited simultaneous calibration interval effectively preserves 95% coverage when a regression coverage factor of 2.02 is applied.
• Accurate statistical methods are key to reliable dose uncertainty estimation in radiation studies.