LBDNet interlaboratory comparison for the dicentric chromosome assay by digitized image analysis applying weighted...

Area of Science:

• Radiological Sciences
• Biomedical Imaging
• Cytogenetics

Background:

• Interlaboratory comparisons are crucial for validating biodosimetry techniques.
• Digital image analysis offers standardized methods for chromosome aberration scoring.
• Uneven expertise among laboratories can impact dose assessment accuracy.

Purpose of the Study:

• To assess the Latin-American Biodosimetry Network (LBDNet) performance in analyzing digitized dicentric chromosome images.
• To evaluate the effectiveness of weighted robust algorithms in standardizing dose estimations across laboratories with varying expertise.
• To compare dose estimations using individual versus common calibration curves.

Main Methods:

• Digitized images of dicentric chromosomes from irradiated blood samples (1.5 Gy and 4 Gy) and a control (0 Gy) were analyzed by LBDNet laboratories.
• Laboratories performed triage scoring, conventional scoring, and dose estimation using their own or a common calibration curve.
• Weighted robust Hampel algorithm and z-score analysis were employed to account for interlaboratory expertise differences.

Main Results:

• Dose estimations showed variability, with some laboratories performing unsatisfactorily at lower doses when using their own calibration curves.
• All doses were satisfactory at 4 Gy.
• Weighted algorithms reduced bias in dose estimation for non-irradiated samples, and a common calibration curve decreased the standard deviation of dose estimates.

Conclusions:

• Interlaboratory comparisons using digitized images are vital, even for non-irradiated samples.
• Weighted robust algorithms are essential for achieving precise biodosimetry outcomes when laboratory expertise varies.
• Standardized digital image analysis and robust statistical methods enhance the reliability of biodosimetry networks.