Establishing a dose-response curve for 60Co γ-ray irradiation-induced micronuclei in human peripheral blood cells using artificial intelligence-based color recognition / 中华放射医学与防护杂志

ABSTRACT

Objective:To identify micronuclei through the cytochalasin B blocking micronucleus method-based assay using scanning microscope, combined with the slide scanning software Metafer 4 and, accordingly, to establish a dose-response relationship between the dose of 60Co γ-rays and the frequency of micronuclei in human peripheral blood lymphocytes using artificial intelligence-based color recognition. Methods:Blood samples were collected from four healthy individuals (two men and two women) and were then exposed to varying doses of 60Co γ-ray radiation (0, 0.25, 0.5, 1, 2, 3, 4, 5 Gy) at a dosage rate of 0.74 Gy/min. Micronucleus slides were prepared as per the GBZ 128-2023 standard. The numbers of binuclear cells and micronuclei were recorded using an artificial intelligence-based color recognition analysis system. The dose-response curve was determined through fitting using the CABAS software. Then, the doses to both independent samples were estimated based on the curve. Results:Within a dose range of 0 to 5 Gy, the fitted micronucleus dose-response curve aligned with a quadratic polynomial model, with a regression equation of y = 0.032 1 D2+ 0.023 7 D+ 0.012 7 ( D denoting the dose, correlation coefficient R2=0.998). The dose estimations from the validation samples closely corresponded to the actual irradiation doses. Conclusions:Establishing the micronucleus dose-response curve provides a feasible method and basis for the rapid and accurate estimation of radiation biological doses in laboratory automation.