Study on the comparison of semi-automatic and manual dicentric detection for biological dosimetry / 中华放射医学与防护杂志

ABSTRACT

Objective:To explore the advantage of dose estimation based on semi-automated dicentric (dic) scoring and the feasibility of reflecting a large scale radiation accident for population clinical triage by mean of the comparison between semi-automatic and manual detection.Methods:Human peripheral blood samples from two healthy volunteers were irradiated by 60Co γ-rays at the doses of 0, 0.5, 1, 2, 3, 4, 5 and 6 Gy with a dose rate of 0.27 Gy/min, and chromosome preparation was carried out using the conventional method. The dic and dic plus ring were analyzed automatically with the DCScore software and manually with the Ikaros software, respectively, in a high-throughput chromosome automatic scanning system. The dose-response curves were fitted with dic or dic plus ring per cell. Twelve standard samples of biodosimetry were used to validate the dose-response curves. Results:The numbers of dic or dic plus ring per cell by semi-automatic or manual detection increased with the exposure doses ( r=0.984, 0.972, 0.972, P<0.01). The yields of semi-automated dic or manually detected dic plus ring were well fitted with the correlation coefficients ( R2=0.998, 0.999, 0.999, P<0.01). When the exposure dose of the standard samples was more than 2 Gy, the relative deviation between actual and predict doses was within 21% using the dose-response curve based on automated dic before human verification and correction of dic (elimination of false positives and inclusion of true positives), and wiht ±10% after manual elimination of false positive dic. Bio-doses estimated from the dic detected by manual scoring were similar to the actual exposure doses with the exception of 0.7 Gy, but the efficiency of semi-automatic analysis of dic was increased by 6-times in bio-dose assessment. Conclusions:The semi-automated dic detection obviously improves the level and efficiency of biodosimetry analysis, and thus can meet the requirements of clinical classification diagnosis of medical emergency response to large-scale nuclear radiation events.