| Abstract: |
Micronuclei, detected through the cytokinesis-block micronucleus assay, are valuable indicators of ionizing radiation exposure, especially in short-term lymphocyte cultures. The peripheral human blood lymphocyte assay is recognized as a prime candidate for automated biodosimetry. In a prior project at the Columbia University Center for Radiological Research, we automated this assay using the 96-well ANSI/SLAS microplate standard format and relied on established biotech robotic systems named Rapid Automated Biodosimetry Tool (RABiT). In this study, we present the application of a similar automated biotech setup at an external high-throughput facility (RABiT-III) to implement the same automated cytokinesis-block micronucleus assay. Specifically, we employed the Agilent BRAVO liquid-handling system and GE IN Cell Analyzer 6000 imaging system in conjunction with the PerkinElmer Columbus image data storage and analysis system. Notably, this analysis system features an embedded PhenoLOGIC machine learning module, simplifying the creation of cell classification algorithms for CBMN assay image analysis and enabling the generation of radiation dose-response curves. This investigation underscores the adaptability of the RABiT-II CBMN protocol to diverse RABiT-III biotech robotic platforms in non-specialized biodosimetry centers. Furthermore, it highlights the advantages of machine learning in rapidly developing algorithms crucial for the high-throughput automated analysis of RABiT-III images. © 2024 by Radiation Research Society. |
