Multiparameter representation and classification of blood cells based on phase map analysis

Abstract

Morphological analysis and feature representation of whole blood cells (white blood cells, red blood cells and platelets) are of great importance in life science and clinical medicine. In order to obtain phases of cells, an experimental platform is used based on the principle of transport of intensity equation (TIE). To obtain the detailed information of the cells, five characteristic parameters of the measured phase including peak-valley difference, phase volume, phase surface area, perimeter and project area are introduced. At the same time, the corresponding characteristic values of erythrocytes, neutrophils and platelets are discussed. Moreover, a phase gradient method is proposed to analyze the interface of the cell and remove the halo in its phase map. Finally, combining with the threshold of the parameters and phase gradient of the cells, an identifying method is established so that three classification of whole blood cells can be realized. Experimental results demonstrate that the proposed method has a better accuracy and good feasibility. It provides a novel method and technical basis for the analysis and classification of blood cells based on quantitative phase imaging.