Analysis of 2D-gel images for detection of protein spots using a novel non-separable wavelet based method. (March 2016)

• Record Type: Journal Article
• Authors: Sengar, Ratnesh Singh; Upadhyay, Ashutosh Kumar; Singh, Manjit; Gadre, Vikram M.
• Is Part Of: Biomedical signal processing and control. Volume 25(2016)
• Page Start: 62
• Page End: 75
• Publication Date: 2016-03
• Subjects: 2D gel image -- Image segmentation -- Non-separable wavelet -- Protein spot -- Quincunx; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2015.10.013
• Languages: English
• ISSNs: 1746-8094

Abstract

Graphical abstract: Highlights: A 2D gel image is modeled using the inter-scale ratio of wavelet coefficients. The image is divided into three distinct parts: spot regions, edges and background. A single threshold is enough to extract the protein spots from the 2D gel image. The proposed method provides more accurate segmentation than available methods. Abstract: 2D-gel electrophoresis (2DGE) is an important technique in proteomics for analyzing the protein expressions. However, the analysis of 2DGE images is still a cumbersome and tedious task. One of the main reasons is the presence of a large amount of the inhomogeneities in the foreground and the background intensities. In this paper, we have proposed a novel approach of segmentation of the protein spots in the non separable wavelet domain. It utilizes the inter-scale relationship among enhanced wavelet coefficients, which can easily distinguish the different features of the image—the interior region of spots, the edges and the background. This technique is based on a single threshold and is independent of the gray value of the image. It copes with the inhomogeneities in the 2DGE images up to a great extent, which is helpful for finding the protein spots accurately. The artifacts are further removed using a non-threshold based method comprising a weighted Gaussian energy distribution model. Experimental results show that our method outperforms the available commercial software and previously reported works.