An automated method for dynamic red blood cell aggregate detection in microfluidic flow.

OBJECTIVE: Red blood cell (RBC) aggregation is a unique phenomenon that occurs when red blood cells are subjected to low shear rates. Little is known about the sizes, shapes and behaviour of aggregates flowing in healthy humans. However, excessive aggregation has been shown to be an indication of pathological conditions. Therefore, characterizing RBC aggregates is important to medical research. The objective of this study was to develop a reliable technique based on image processing to assess and characterize human RBC aggregation subjected to controlled and measurable shear rates in a two-fluid flow microfluidic shearing system.

APPROACH: Images of RBC suspensions at [Formula: see text], [Formula: see text] and [Formula: see text] entrained by a phosphate buffered saline solution in a PDMS microchannel were captured with a high speed camera. An algorithm for processing the RBC aggregate images is presented and validated (1) on a sample of known diameter hollow glass microspheres and (2) by comparing RBC aggregate size results with those of an ImageJ image processing technique and those obtained by manual detection by two independent researchers.

MAIN RESULTS: The proposed image processing algorithm provides a very good agreement with the manufacturer data for the glass microspheres. It also performs well on the RBC suspension images, with errors of 2-4 [Formula: see text] with respect to the manual results.

SIGNIFICANCE: The proposed automated method for RBC aggregate detection is found to be reliable and fairly accurate and will serve researchers and, perhaps in the future, clinicians to assess healthy and pathological RBC aggregation under flowing conditions.