PROJECT TITLE :

Cell segmentation in digital holographic images - 2016

ABSTRACT:

Digital Holographic Microscopy (DHM) is turning into recently terribly standard for cell imaging. The most advantage of digital holographic microscopy over classical microscopy techniques is that it will not only offer the projected image of the article however conjointly provides three dimensional info of the article's optical thickness. DHM technology could be the core of a label-free imaging for hematology applications. In an ideal framework, a blood sample will be imaged using DHM, machine learning approaches will be used for the cell extraction, differentiation and consequently computing all the relevant blood statistics like the Mean Corpuscular Volume (MCV), the Red Blood Cell (RBC) count, Red Blood Cell Distribution Width (RDW). The most important element in such a framework is correct extraction of the cells. This paper presents a completely unique approach to cell segmentation in which a probabilistic boosting tree classifier is trained to detect the centers of the cells using Haar-Options. The detected cell centers are used to trigger a marker-controlled power watershed segmentation to compute the cell boundaries. Additionally, we gift a comprehensive analysis of segmentation strategies for cell extraction in digital holographic pictures.