Microscopic image segmentation based on pixel classification and dimensionality reduction

Abstract

Pathological image analysis plays a significant role in effective disease diagnostics. In this article, a tool for diagnosis assistance by automatic segmentation of bone marrow images is introduced. The aim of our segmentation is to demarcate cell's component: nucleus, cytoplasm, red cells, and background. Different color spaces were used to extract color's features to profit of their complementarity. We introduce several dimensionality reduction techniques. These techniques are exemplified on a support vector machine pixel‐based bone marrow image segmentation problem in which it is shown that it may give significant improvement in segmentation accuracy and time consuming. © 2013 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 23, 22–28, 2013

Automated identification of plasma cell in bone marrow images

“…These results are significantly better compared to our previous work in terms of segmentation quality and in terms of time and computing [7], [8].…”

“…The adopted model is guided by the two strategies offered by information fusion, i.e. classifying separately the data from different sources(color spaces) then merging decisions, or combining these data to classify them [7]. The second [8] deals with new segmentation framework based on evidence theory, called ESA (Evidential Segmentation Algorithm).…”

Adaptable image segmentation via simple pixel classification

“…Pixel classification as a path to image segmentation has excellent examples including other works –. This is very closely related to superpixel‐ and tile‐based image segmentation as deployed by other works –.…”

“…Pixel classification as a path to image segmentation has excellent examples including other works. [5][6][7][8][9] This is very closely related to superpixel-and tile-based image segmentation as deployed by other works. [10][11][12] Another approach to easing the complexity problem borrows from the natural idea of foveation to reduce the amount of information available about a pixel as one moves farther away from that pixel.…”

Data mining system for tree and network topologies visualised in medical images

“…For example, topological properties of the pulmonary airway tree are associated to the diagnosis and treatment of lung disorders and are considered as biomarkers in physiological studies of pulmonary function , Feragen et al 2013. Analysis of vessel tortuosity has been proposed as a marker of malignancy of brain lesions [Bullitt et al 2004, Bullitt et al 2005. In digital fundus photography, measures of retinal topography are related to atherosclerosis risks [Hubbard et al 1999] and arteriolar attenuation affected by diabetic retinopathy [Quigley and Cohen 1999].…”