A Novel Automated Slide-Based Technology for Visualization, Counting, and Characterization of the Formed Elements of Blood: A Proof of Concept Study.

ABSTRACT

CONTEXT - A novel automated slide-based approach to the complete blood count and white blood cell differential count is introduced. OBJECTIVE: - To present proof of concept for an image-based approach to complete blood count, based on a new slide preparation technique. A preliminary data comparison with the current flow-based technology is shown. DESIGN: - A prototype instrument uses a proprietary method and technology to deposit a precise volume of undiluted peripheral whole blood in a monolayer onto a glass microscope slide so that every cell can be distinguished, counted, and imaged. The slide is stained, and then multispectral image analysis is used to measure the complete blood count parameters. Images from a 600-cell white blood cell differential count, as well as 5000 red blood cells and a variable number of platelets, that are present in 600 high-power fields are made available for a technologist to view on a computer screen. An initial comparison of the basic complete blood count parameters was performed, comparing 1857 specimens on both the new instrument and a flow-based hematology analyzer. RESULTS: - Excellent correlations were obtained between the prototype instrument and a flow-based system. The primary parameters of white blood cell, red blood cell, and platelet counts resulted in correlation coefficients (r) of 0.99, 0.99, and 0.98, respectively. Other indices included hemoglobin (r = 0.99), hematocrit (r = 0.99), mean cellular volume (r = 0.90), mean corpuscular hemoglobin (r = 0.97), and mean platelet volume (r = 0.87). For the automated white blood cell differential counts, r values were calculated for neutrophils (r = 0.98), lymphocytes (r = 0.97), monocytes (r = 0.76), eosinophils (r = 0.96), and basophils (r = 0.63). CONCLUSIONS: - Quantitative results for components of the complete blood count and automated white blood cell differential count can be developed by image analysis of a monolayer preparation of a known volume of peripheral blood.