Vol. 43, Issue 3, pp. 505-514

Abstract

The digital holographic phase microscopy (DHPM) technique which has been proposed for cellular morphology and dynamic analysis yielded highly desirable results. However, for nucleated cells (especially white blood cells (WBCs)), their submicroscopic structure has not yet been deconstructed through a phase unwrapping method due to the heterogeneity of an internal phase. By analyzing the phase heterogeneity of subclasses of WBCs, the typical phase models of them are built first in this paper; using the simulation method, the wrapped phase distributions of these models are obtained. However, by optimizing the wrapped phase maps and analyzing the relationships between them and typical blood cells, their features are selected and extracted. Then the models built are sorted out from each other successfully without unwrapping via analyzing these extracted features, which provides a valuable approach and technological base for the classification and identification of blood cells.