| 英文摘要 |
In recent years, fully automated digital image analysis technology has the potential to gradually replace most manual microscopic morphological examinations. High-quality smears are essential for automated digital image analysis to accurately identify and classify cells. After evaluation, it was found that cell distribution significantly affects the analysis. Therefore, based on the spreading speed, angle, and drop size, three smear preparation protocols (A, B, and C) were set up to compare the classification of normal samples by the analyzers. The consistency and correlation of the five types of white blood cell differentials between the XN-10 analyzer and manual review by the DI-60 was compared using the Intraclass Correlation Coefficient (ICC) and linear regression, and their differences were analyzed using paired samples t-tests. The results showed that the laboratory chose the best protocol, C, as the smear preparation setting to improve consistency and thereby enhance the quality of hematological analysis. |