, 1997 ), which detect cell boundaries by minimizing a predefined energy functional, can result in poor boundary detection because they use local optimization algorithms that only guarantee to find a local minimum or use the gradient vector field of the image to decode the boundary information ( Fig. Similarly, popular deformable model approaches such as geodesic active contours ( Caselles et al. , 2005 ) or the watershed transform ( Meyer, 1994 ) can result in inaccurate cell boundaries by misspecifications of the cell region to be divided or by oversegmentation ( Fig. In particular, region accumulation approaches such as Voronoi-based methods ( Jones et al. Most current methods use a few basic algorithms for cell segmentation: intensity thresholding, filtering, morphological operations, region accumulation or deformable models ( Meijering, 2012 ). Finally, different experimental configurations such as cell types or imaging protocols generate images with greatly varying morphological or intensity characteristics. Third, growing cell populations usually result in dense cell regions this makes it hard to assign image features to the correct cell, especially among sets of spatially close cells. Second, cell compartmentalization as well as intra- and inter-cell variability induces non-homogeneous marker distributions within and across cells, leading to undesirable image features such as intensity gradients. The objects have heterogeneous shapes that are typically subject to dynamic changes mathematical shape models are therefore nearly impossible to define. First, segmenting cellular images requires the identification of multiple objects in the image. von Meijering: ‘Rather than converging to a robust, unified solution, it thus seems that the field is diverging, and by now almost as many cell segmentation methods have been developed as there exist cell analysis problems …’ ( Meijering, 2012 ).Ĭell segmentation is challenging ( Peng, 2008 ) for many reasons. For good segmentation results, current approaches are typically applicable to narrowly defined image acquisition protocols ( Gordon et al. , 2012 ), the development of cell segmentation algorithms is lagging behind. Although there is a rapid development of imaging hardware and image analysis software platforms ( Eliceiri et al. Accurate quantification of such features critically depends on the spatial detection of the cells in the image, that is, on cell segmentation ( Li et al. Nowadays, optical microscopy is widely used to quantify single-cell features, such as cell size or intracellular densities of fluorescent markers. Ĭontact : or information: Supplementary data are available at Bioinformatics online. As a proof of concept, we monitor the internalization of green fluorescent protein-tagged plasma membrane transporters in single yeast cells.Īvailability and implementation : Matlab code and examples are available at. ![]() ![]() In quantitative benchmarks and comparisons with established methods on synthetic and real images, we demonstrate significantly improved segmentation performance despite cell-shape irregularity, cell-to-cell variability and image noise. The method accurately segments images of various cell types grown in dense cultures that are acquired with different microscopy techniques. Graph cuts account for the information of the cell boundaries through directional cross-correlations, and they automatically incorporate spatial constraints. Results: We present an automatic cell segmentation method that decodes the information across the cell membrane and guarantees optimal detection of the cell boundaries on a per-cell basis. This has not yet been algorithmically exploited to establish more general segmentation methods. However, many microscopy protocols can generate images with characteristic intensity profiles at the cell membrane. ![]() In addition, most current segmentation algorithms rely on a few basic approaches that use the gradient field of the image to detect cell boundaries. Although a plethora of segmentation methods exists, accurate segmentation is challenging and usually requires problem-specific tailoring of algorithms. Motivation: Identifying cells in an image (cell segmentation) is essential for quantitative single-cell biology via optical microscopy.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |