Greedy Approach In Analyzing Multi-Clustered Cell Nuclei On Pap Smear Images

The Advent of Cytological driven cancerdiagnosis I widely accepted with the constanteffort of Dr.Papincolaou's research worksince 1943. Demanding need to automate thisimage processing approach for precancerousdetection has gone through multiple phases inthe last decade [1] [2]. This paper proposes aGreedy Algorithm approach for an automatedprocessing of the Pap smear images inidentifying the malignant cells using chromaticbehavior of the stained images. The uniquecontribution of this approach is to process theimages with the less number of cycles anddiscriminate normal cells from cervical cancercells thereby increase the efficiency of thesystem while not overlooking the accuracy. Ascompared to the classical approaches appliedfor the same medical challenge 0.07% to0.03% false positive result is proven by thisapproach. This paper also highlights the needfor overlapped nucleus processing [3] usingthe same approach that helps in Ever since the establishment of cell theory in the beginning 19th century, which recognized the cell as the fundamental building unit of life, biologists have sought to detail the underlying principles. Significant discoveries were made over the course of many decades of research [4]. The longing desire to attain a complete understanding of the cellular mechanism and how to manipulate them is the constant effort spent till this day by many researchers across the globe. Image Segmentation is often considered as the keystone of image analysis process. Specifically since cellular morphology is an important phenotypic feature that is symptomatic of the physiological state of a cell, and the since the cell contour is often required for subsequent analysis of intracellular process (Zooming in or zooming out) the problem of cell segmentation has received increasing attention in the past [4] detecting thehidden cells during automated image analysis. There has been similar work carried out bymany forerunners but this paper uniquely address how chromatic behavior of the imagecan be used in identifying cell differences.