Automated digital cell morphology identification system (CellaVision DM96) is very useful for leukocyte differentials in specimens with qualitative or quantitative abnormalities

Int J Lab Hematol. 2013 Oct;35(5):517-27. doi: 10.1111/ijlh.12044. Epub 2013 Jan 3.

Abstract

Introduction: The CellaVision DM96 system (CellaVision AB, Lund, Sweden) was developed as one of the automated digital cell morphology analyzer for determining leukocyte differential counts in peripheral blood smears (PBS) and we evaluated this system.

Methods: A total of 308 PB samples with abnormalities were analyzed in this study. For each sample, manual differential counts were performed by two independent technologists, and the CellaVision DM96 system was applied in duplicate. Correlations between the two methods and ability of this system to identify six abnormalities were assessed.

Results: The correlation coefficients between two methods were consistently high, ranged from 0.864 to 0.992. The sensitivity, specificity, positive predictive value, negative predictive values of this system for the identification of abnormalities were consistently high, especially for blasts (98.2%, 99.2%, 96.6%, 99.6%). When the instrument was ordered to count 300 or 500 cells from the operator, better performance was demonstrated than 100 cells in the leukopenic samples by sacrificing only 40 s/slide in average.

Conclusions: The CellaVision DM96 system is useful in the clinical laboratory providing comparative accuracy compared with manual counts in samples with abnormalities. In leukopenic samples, report quality can be improved by ordering to count 300 or 500 cells from the operator without severe prolongation of turnaround time.

Keywords: Abnormalities; automated digital cell morphology identification system; automatic hematology analyzer; leukocyte differentials; leukopenia.

MeSH terms

  • Humans
  • Leukocyte Count / instrumentation
  • Leukocyte Count / methods*
  • Leukocyte Count / standards
  • Leukocytes / cytology*
  • Leukocytes / pathology*
  • Reproducibility of Results
  • Sensitivity and Specificity