Experience with CellaVision DM96 for peripheral blood differentials in a large multi-center academic hospital system

Abstract

Context and Aims:Rapid, accurate peripheral blood differentials are essential to maintain standards of patient care. CellaVision DM96 (CellaVision AB, Lund, Sweden) (CV) is an automated digital morphology and informatics system used to locate, pre-classify, store and transmit images of platelets, red and white blood cells to a trained technologist who confirms or edits CV cell classification. We assessed our experience with CV by evaluating sensitivity, specificity, positive predictive value and negative predictive value for CV in three different patient populations.Materials and Methods:We analyzed classification accuracy of CV for white blood cells, erythroblasts, platelets and artefacts over six months for three different university hospitals using CV.Results:CV classified 211,218 events for the adult cancer center; 51,699 events for the adult general hospital; and 8,009 events for the children's hospital with accuracy of CV being 93%, 87.3% and 95.4% respectively. Sensitivity and positive predictive value were <80% for immature granulocytes (band neutrophil, promyelocyte, myelocyte and metamyelocytes) (differences usually within one stage of maturation). Cell types comprising a lower frequency of the total events, including blasts, showed lower accuracy at some sites.Conclusions:The reduced immature granulocyte classification accuracy may be due in part to the subjectivity in classification of these cells, length of experience with the system and individual expertise of the technologist. Cells with low sensitivity and positive predictive value comprised a minority of the cells and should not significantly affect the technologist re-classification time. CV serves as a clinically useful instrument in performance of peripheral blood differentials.

Performance of CellaVision DM96 in leukocyte classification

“…It has previously been shown that the DM96 is faster than manual smear examination an aspect confirmed in our internal departmental validation studies. [9–121415] This is of particular importance as labor is one of the major expenditures in the laboratory. In addition, digitized images provide many advantages compared to manual slide microscopy.…”

Evaluation of the accuracy of the CellaVision™ DM96 in a high HIV-prevalence population in South Africa

“…Our misclassification rate was higher than that described previously by Rollins-Raval, Raval and Contis, who assessed the performance of the DM96 over a six-month period at three separate sites. 14 The number of unidentified cells in that study was ~1% and the number of misclassified cells ranged from 4.6% – 12.7%. The higher misclassification rate in our study is likely a product of the large number of specimens containing malignant cells and selection of a large number of samples with extreme WCCs for the purpose of evaluating the analyser.…”

Deep Learning Approaches and Applications in Toxicologic Histopathology: Current Status and Future Perspectives

“…[ 65 105 106 107 ] In human and veterinary diagnostic cytopathology, proprietary, neural-network-based pre-classification systems arewidely used, particularly for hematology applications. [ 108 109 ] Most recently, two AI-based decision support tools in CDP have gained FDA Breakthrough Device designation: Paige. AI in 2019 for cancer histology[ 5 ] and 4D Q-plasia OncoReader Breast in 2020 for breast cancer histology.…”