Integrating the gamma-H2AX assay into an open-source software for radiation dosimetry

In biological dosimetry, the dicentric assay has been the most commonly used method for retrospective estimation of radiation dose following suspected exposure to ionizing radiation. The open-source software Biodose Tools (Hernández et al, 2023) supports this approach by providing model fitting and dose estimation functionalities. While the dicentric assay remains the gold standard, gamma-H2AX foci offer the advantage of faster assessment in emergencies and have been developed in recent years. In this presentation we report on recent work regarding the inclusion of the gamma-H2AX foci assay into Biodose Tools, covering both the tasks of estimation of the dose-response relationship from calibration data (based on quasi-Poisson models), as well as the estimation of the contracted radiation dose from an exposed patient sample (of known exposure time) via an inverse regression routine. Uncertainty quantification is possible via the delta-method. Statistically, a particular challenge resides in estimating, and accounting for, (over-)dispersion at various stages along this pipeline. We will also give some attention to the production of a user-friendly R Shiny interface which is compatible with other Biodose Tools functionalities.We further explore an extension of the provided functionalities which accounts for a potentially unknown time of exposure as proposed by Cai et al. (2024). The methodology makes use of the nature of the foci decay. Given two biomarker measurements with known time distance but unknown time of exposure, this approach yields two complex equations which can be solved (for dose and time) with numerical solvers. We explore the use of the R solver nleqslv() in connection with a large gamma-H2AX dataset, discussing in particular the challenges arising due to the sensitivity of the routine to the choice of starting values, which has particularly implications for uncertainty quantification.Cai Y, Einbeck J, Barnard S, Ainsbury E (2024). Estimating Dose and Time of Exposure from a Protein-Based Radiation Biomarker. In: Einbeck J, Maeng H, Ogundimu E, Perrakis K (eds) Developments in Statistical Modelling. IWSM 2024. Contributions to Statistics. Springer, Cham. https://doi.org/10.1007/978-3-031-65723-8_37Hernández A, Endesfelder D, Einbeck J, Puig P, Benadjaoud MA, Higueras M, Ainsbury E, Gruel G, Oestreicher U, Barrios L, Barquinero JF. (2023). Biodose Tools: an R shiny application for biological dosimetry. Int J Radiat Biol. 99(9):1378-1390. doi: 10.1080/09553002.2023.2176564.