Mumford-Shah Model-Based Image Inpainting for Enhancing Protein Spot Detection in Two-Dimensional Gel Electrophoresis Images

Many software programs are made for processing and analyzing two-dimensional gel electrophoresis (2-DGE) images. However, the anomalies existing in these 2d gel images make the reliable system needed to achieve for 2-DGE image processing and analysis still difficult to realize. In this paper, we propose a new 2-DGE images preprocessing technique. Our new technique depends on image inpainting using Mumford Shah Euler Lagrange to erase vertical and horizontal streaking from these images. Also, we compare the analysis of 2-DGE images inpainted by our proposed technique with non-inpainted images using the known commercial software Delta2D, and we compute the F-measure in both cases for three benchmark 2-DGE images. The results show a significant improvement when comparing analysis of inpainted images with non-inpainted images in subjective and objective ways.