Hierarchical Partial Matching and Segmentation of Interacting Cells

Abstract

Abstract. We propose a method that automatically tracks and segments living cells in phase-contrast image sequences, especially for cells that deform and interact with each other or clutter. We formulate the problem as a many-to-one elastic partial matching problem between closed curves. We introduce Double Cyclic Dynamic Time Warping for the scenario where a collision event yields a single boundary that encloses multiple touching cells and that needs to be cut into separate cell boundaries. The resulting individual boundaries may consist of segments to be connected to produce closed curves that match well with the individual cell boundaries before the collision event. We show how to convert this partial-curve matching problem into a shortest path problem that we then solve efficiently by reusing the computed shortest path tree. We also use our shortest path algorithm to fill the gaps between the segments of the target curves. Quantitative results demonstrate the benefit of our method by showing maintained accurate recognition of individual cell boundaries across 8068 images containing multiple cell interactions.

Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images: A Comprehensive Review

“…For an image, the segmentation procedure begins with extracting cell boundaries using color GVF snake [196], then constructing a graph with nodes and edges corresponding to concave points and inner edges inside cells, respectively, and finally recursively searching for optimal shortest paths in the graph to separate touching cells. Wu et al [294] have segmented interacting fibroblast cells in phase-contrast microscopy images by seeking the shortest path in a graph. With two endpoints of a gap, the graph is constructed by using the pixels between these two endpoints as nodes and the image gradient as the edge weight.…”

MPM: Joint Representation of Motion and Position Map for Cell Tracking

“…Moreover, many optimization methods have been proposed to associate the detected cells. They maximize an association score function, including linear programming for frame-by-frame association [21,5,6,46,42] and graph-based optimization for global data association [8,36,39,15]. These methods basically use the proximity and shape similarity of cells for making hand-crafted association scores.…”

High-throughput histopathological image analysis via robust cell segmentation and hashing

“…Lucchi et al (Lucchi et al, 2010) exploited a mincut-maxflow algorithm to partition the superpixel based graph, Bernardis and Yu (Bernardis and Yu, 2010) segmented out individual cells based on the normalized cuts (Shi and Malik, 2000), and Zhang et al (Zhang et al, 2014a) employed a correlation clustering method to achieve superpixel graph partition. Some other graph based methods can be found in (Al-Kofahi et al, 2010; Nath et al, 2006; Faustino et al, 2009; Chen et al, 2008; Wu et al, 2012; Yu et al, 2010; Janowczyk et al, 2012; Lou et al, 2012). Although effcient graph-based segmentation algorithm (Felzenszwalb and Huttenlocher, 2004) is proposed, generally graph partition methods exhibit high time cost, which limits their applications in real cell segmentation.…”