Artificial Intelligence in Digital Holographic Imaging

18Automated Tracking of Temporal Displacements of Red Blood Cells

18.1 Introduction

RBCs, particularly spherocytes, may exhibit temporal displacement or movement when their membranes are loosely attached to their substrates during sedimentation on a glass surface. In addition, RBCs investigated with time‐lapse DHM may undergo lateral displacement due to microscope drift. Consequently, it is inevitable to develop a tracking algorithm to automatically investigate RBCs with temporal displacement or movement over time even though these RBCs are attached to a coverslip (they show temporal displacement when their membranes are loosely attached to a glass surface or due to microscope drift). By using a tracking scheme, it is possible to quantitatively analyze temporal displacement and 3D morphology of RBCs. This scheme may also be used in other studies like analyses of RBC fluctuation since it can localize the same RBC in time‐lapse sequences. In this chapter, we will introduce a tracking algorithm to measure temporal movements of RBCs along x and y axes. We will focus on RBCs attached to a coverslip with 2D movements.

For cell tracking, there are several algorithms based on segmentation and tracking [1, 2]. These approaches first segment target objects and then establish the association of segmented cells between successive frames. Segmentation‐based tracking algorithms need target objects to have a sharp border and the same object in adjacent frames to share an overlapping area. ...