Summary Following a large scale radiological or nuclear event, hundreds of thousands of people may be exposed to ionizing radiation/s and require subsequent dose-dependent medical management. It will be crucial to collect and analyze human biofluids (such as blood, urine, saliva) as soon as possible within the first week for accurate dose prediction and early triage decision. There is a need for FDA-approved in vitro diagnostic high-throughput biodosimetry devices with the ability to determine past radiation exposure with precision and accuracy. At the Center for High Throughput Radiation Biodosimetry, the Columbia University Center for Medical Countermeasures against Radiation (CMCR), we have developed FAST-DOSE (Fluorescent Automated Screening Tool for Dosimetry) assay system, to measure radiation-responsive proteins in human peripheral blood samples for retrospective estimation of radiation dose. The protein panel also includes biomarkers for blood leukocyte subtypes to reflect hematological sensitivity and injury. The FAST- DOSE assay system is intended as an in vitro diagnostic device (IVD) as defined by 21 CFR 809.3. The platform uses a commercial imaging flow cytometry system (ImageStream®X) and associated Image Data Exploration and Analysis Software (IDEAS®) to rapidly quantify changes in biomarker expression levels within specific cellular structures using fluorescent imagery and algorithms for estimation of absorbed dose. The studies planned here are designed to develop and optimize our FAST-DOSE assay system to accurately estimate absorbed dose and assess hematopoietic injury in human lymphocytes after ionizing irradiation. The first objective is to build on our current biomarker validation data for early engagement with the FDA via the pre-submission process. We have used the human ex vivo model and humanized mouse (Hu-NSG) and non- human primate (NHP) models to validate biomarker expression and radiosensitivity in blood leukocytes after acute ionizing radiation exposure. The Specific Aims proposed here are designed to: optimize the assay protocol and identify biomarker dose/time kinetics for accurate dose predictions in vitro and test 1) inter-donor variation, 2) intra-donor variation and 3) inter-laboratory variability (Aim 1); test the effect of specific confounders: age and sex, inheritance with germline BRCA1/2 pathogenic variant, and inflammation and trauma on the biomarker response, before and after irradiation (Aim 2); measure biomarker levels and time kinetics in vivo and correlate with hematopoietic injury, based on peripheral blood leukocyte counts, and stem and progenitor cell levels in the bone marrow of Hu-NSG mice (Aim 3) and, develop mathematical models (using machine learning and regression techniques) to select the best FAST-DOSE biomarkers and their combinations for generating dose predictions based on the ex vivo and in vivo dose response of these biomarkers (Aim 4). Our vision for future development is to develop ...