Fourier transform infrared biodosimetry on skin

PROJECT SUMMARY We aim to develop a non-invasive, painless biodosimeter for use during triage after a radiological emergency. Our preliminary data showed that Fourier transform infrared attenuated total reflection (FTIR-ATR) imaging coupled to statistical machine learning models is capable of distinguishing irradiated from control mice at doses as low as 0.1 Gy for as long as 90 days after radiation exposure. To further develop this signature, and to determine the ability to discriminate between radiation doses, we will first increase our power of detection by expanding our studies to human skin explants. Statistical machine learning models will be used to develop an FTIR-based signature that can distinguish irradiated from unirradiated animals and accurately determine radiation dose and time since exposure for triage. We will then use ambient infrared laser ablation mass spectrometry to identify the biomolecules that underly the radiation-specific response signature in the skin and address confounders. This novel, non-invasive procedure will lay the groundwork for future applications of deployable imaging devices for biodosimetry at population scales – a valuable tool for radiobiology, epidemiology, and monitoring. Further, because only hundreds of samples were required to learn highly discriminative signatures in our pilot study, this study will contribute to developing human-relevant diagnostic capabilities.