| Abstract [eng] |
In case of large scale radiation accident, the absorbed dose of potentially exposed individuals should be estimatet as quickly as possible to differentiate individuals with low exposure dose from those requiring urgent medical intervention. Even though the manual dicentric assay is the most validated and widely used assay in biological dosimetry for individual dose estimation, it is laborious and time consuming. Therefore, to increase the throughput of the dicentric assay, the automatic dicentric assay was developed and validated as an accurate method for preliminary dose estimation in recent years. The automatic dicentric assay involve 3 steps: the automatic metaphase finding (Msearch) which is performed using a 10x objective, capturing images of high resolution at 63x magnification (Autocap) and automatic detection of dicentric chromosomes which is validated by a human scorer. In this study, the automatic dicentrics scoring with 63x objective was compared to scoring with 40x in an attempt to increase the throughput of the automatic dicentric assay. Whole blood samples were irradiated with low LET X-Ray at doses between 0 and 5 Gy. After slides preparation, dicentric scoring was performed both automatically (with 63x and 40x objective) and manually (only for doses of 1 and 4 Gy). The time taken to perform each step of the automatic scoring was recorded for analysis of both possibilities. The results indicate that analysis using the 40x objectives lead to a 45% reduction in the total time needed for dicentrics scoring. In addition, the dose effect curves established in the case of the usage of both objectives merge at low doses; however, the dose effect curve for the 40x objective is significantly lower than for the 63x magnification at doses higher than 2 Gy. In addition, the rate of false positive and false negative evaluations was comparable in both cases. The mean false negative rate was 47.7% for scoring with the 40 x objective as compared to 55% for scoring with the 63x objective. Moreover, the same mean false - positve rate was obtained for both objectives (0.11% with 40x objective and 0.1% with 63x objective). In summary, our results show that automatic scoring with 40x objective can support a large-scale sample analysis and decrease the turnaround time of the dicentric assay while maintaining comparable dicentric detection efficiency to the scoring done with the 63x objective. . |
