Identification and Counting of Erythrocytes of Native Human Donor Blood by Digital Optical Microscopy Using Spectral Filtered Illumination

We studied the possibility of identification and counting of erythrocytes of native donor blood, based on static, nonflow digital optical microscopy. The object of the study was whole donor blood diluted with saline and placed in the Goryaev chamber. The sample was examined in transmitted light using a Lumam R-8 digital optical microscope equipped with a Basler acA1920-40um camera. To identify erythrocytes by spectral characteristics, two sets of micrographs, 20 pieces each, were obtained. In one set, there was no optical filter in the illumination channel of the microscope, and in another set, an interference filter at a wavelength of 420 nm with a bandwidth of 10 nm was used. The characteristics of the interference filter were selected from considerations of proximity to the Soret band of erythrocyte hemoglobin to obtain the best contrast for photographs. Using the OpenCV software toolkit, we developed a procedure for automated analysis of micrographs, which makes it possible to recognize erythrocytes and count them. The results of the programmed counting of erythrocytes were compared with the number of erythrocytes counted manually. With the proposed computerized approach, the proportion of recognized erythrocytes averages 97-98%.