Automated Reticulocyte Counting in Peripheral Blood Smears

Abstract

The results of reticulocyte counting obtained using the Vision Hema^® Reticulocytes (Vision Hema^® RET) software are presented and compared to the results obtained by manual reticulocyte counting. Blood samples with pronounced reticulocytosis (main group, n = 15) and samples with normal cell count (control group, n = 15) are analyzed. Use of a scanner analyzer allows the reticulocyte counting time in the samples of the main group to be reduced by 90.5% (from 202.3 ± 44.7 to 106.3 ± 18.3 s); in the control group it is reduced by 56% (from 155.1 ± 38.0 to 99.9 ± 16.8 s). The results of reticulocyte counting using the Vision Hema^® RET software do not differ from the light microscopy data in the main group (p = 0.211) and in the control group (p = 0.53). The Spearman’s correlation coefficient R is 0.977. The results of evaluation of the within-run precision (CV_WR) demonstrate high reproducibility of reticulocyte count as compared to the manual method in cases of both normal (9.45% against 20.5%) and high (4.43% against 8.63%) reticulocyte counts. The Vision Hema^® RET soft-ware is recommended for use in clinical diagnostic laboratories.